摘要:

OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc.it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,

ISSN:0003-2654    

DOI:10.1039/AN9376200339

出版商:RSC    

年代:1937

数据来源: RSC

摘要:

340 BARGER THE ALKALOIDS OF ERGOT The Alkaloids of Ergot BY GEORGE BARGER HON. D.Sc. HON. M.D. LL.D. F.R.S. (Address given at the A.nutua1 General Meeting March 5 1937) LIKE many other drugs ergot owes its therapeutic and toxic properties to alkaloids. It yields four pairs of interconvertible isomerides each pair consisting of a laevo-rotatory alkaloid having a powerful pharmacological action and a strongly dextro-rotatory one with but little pharmacological activity. The following table gives for each alkaloid the year of its discovery and the value of [a],,,. (after Smith and Timmis): Pharmacologically potent Almost inert Ergotoxine - 226' "T Ergotinine + 466' C3,Hsg0,N5 (1906) (1875) It $-Ergotinine + 513" f 1931) Ergotamine (1920) Ergosine (1936) Ergometrine (1935) - 181" - 193' - 16" ~- I Ergotaminine + 450" Ergosinine + 522' Ergometrinine + 596" (1920) (1936) (1936) Associated with the first pair is an additional alkaloid of high dextrorotation (Smith and TimmS5) so that there is in reality a triplet ; perhaps a third isomeride may later be added to the other pairs also.Ergoclavine ( [a] + 124") is according to Kiis~ner,4~ a molecular compound of an alkaloid (a],- 149" resembling ergotoxine and one with [a],+41Oo, resembling ergotinine for which the name ergoclavinine is suggested ; this homo-geneous well-crystallised substance is doubtless an equimolecular compound of ergosine [a]",0-161° and ergosinine [a] + 420"; this view is supported by the products of its hydrolysis. Sensibamine is according to S t 0 1 1 ~ ~ ~ ~ ~ a similar but much less stable molecular compound of ergotamine and ergotaminine ; on mere solution in alcohol or acetone the latter alkaloid crystallises out.The mixture can also be separated by chromatographic adsorption. All the above-mentioned alkaloids have been crystallised. Simple bases insoluble in ether such as ergothioneine uracil guanosine and amines derived from amino acids have been excluded from this review. The four pairs of isomeric alkaloids are particularly well characterised and, in spite of their complexity the molecular formulae have in all cases been established with certainty and confirmed by hydrolysis. This is due largely to the admirable work during the last few years of Jacobs and in the United States and of Smith and TimmissQ-72 in this country.As an example ergotinine yields with aqueous alkali one equivalent of ammonia (S~ltys'~) isobutyryl formic acid (CH,),CH.CO.COOH and one equivalent of lysergic acid C,,H,,O,N (Jacobs and Craig31$32) or with alcoholic sodium hydroxide these fragments may remain combined as the amide of lysergic acid or ergine C,&f170N3 (Smith and Timmis66$6*). Acid hydrolysis destroys lysergic acid (which is an indole derivative BARGER THE ALKALOIDS OF ERGOT 341 and yields I-phenylalanine and d-proline. following equation : The exact arrangement of the various fission products is not known. The lysergic acid is certainly joined to the ammonia (because of the formation of ergine); the isobutyryl formic acid is also joined by its carboxyl to an amino group for on pyrolysis isobutyryl formamide sublimes (Barger and Ewinss).The nitrogen atom in the latter amide need not however be that giving rise to ammonia but may be the nitrogen atom of phenylalanine; the two amino acids are also joined together for their dipeptide has been isolated as a result of moderate hydrolysis. The following formulae for ergotinine have been suggested by Turners6: It is thus possible to deduce the C35Hs905N5 + 4H2O = &Hi,OzN + NH + CSHBO~ + C&,O& + C,HgO&. OH C15H15N .CO .NH .CO .C. CH I c I CH,-CH O.NH c,,H,,N,.co .NH .CO.~H d ~ v N T1 I CO.CH.CH,.C,H, NH.CO.CO. 1 c.<“”” CH3 The only difference between ergotoxine-ergotinine and ergotamine-ergotaminine is that the latter pair give rise to pyruvic instead of isobutyrylformic (= dimethyl pyruvic) acid.Since the rest of the large molecule is the same the great similarity in the pharmacological action of ergotoxine and ergotamine becomes intelligible. The third pair of alkaloids also yields pyruvic acid but here the phenylalanine is replaced by leucine. (These same two products but not d-proline were obtained from ergoclavine by Jacobs and Craig$7s40.) Ergosine and ergoclavine have a pharmacological action closely resembling that of ergotamine and ergotoxine but the presence of leucine instead of phenylalanine gives rise to some quantitative differences as between ergosine on the one hand and ergotamine and ergotoxine on the other. The latter two and probably all three can produce gangrene; all three inhibit the action of adrenaline.In this respect they differ entirely from the active alkaloid of the last pair ergometrine has no such specific pharmacological actions and for this reason escaped the attention of pharmacologists. It was only discovered by exact clinical experiments for its outstanding property is rapidly to bring on powerful contractions of the puerperal (human) uterus. This is the only specific biological test for the substance and ergometrine is par excellence the long-sought therapeutic principle of ergot. The great biological difference from the other alkaloids (which indeed also cause contraction of the uterus but more slowly) is due to a considerable difference in structure and to a much smaller molecular weight whic 342 BARGER THE ALKALOIDS OF ERGOT favours rapidity of absorption ; ergometrine gives on hydrolysis two products only lysergic acid and a simple amine a-hydro~y-/3-arninopropane~~ : The fission products of the various ergot alkaloids are recorded in the following table Ergotoxine Ergotamine Ergosine Ergometrine &,HzaO,Ns + KOH = C;,HI,Nz.COOK + NH,.CH(CH,) .CH,OH.Ergotinine Ergotaminine Ergosinine Ergometrinine t 4 - Lysergic acid * . + + + + Ammonia . . + + Hydroxyisopropylamine . . - - + d-Proline . . * . + + + + E-Phenylalanine . . - . + I-Leucine . . * . Dimethylpyruvic acid . . + - -Pyruvic acid . . + + --- ---- -- -It is of interest to note that the complicated alkaloids of ergot differ rather con-siderably from the alkaloids of the higher plants; the first three pairs are akin to First mention of ergot and of its therapeutic properties in Adam Lonicer's Kreuterbuch Frankfurt on Main 1582 polypeptides and contain indeed two amino acids apiece.In respect of its alkaloids as in respect of many simpler constituents ergot presents points of unusual biochemical interest. The phenylalanine of ergotoxine and ergotamine, and the leucine of ergosine are optically identical with fission products of protein, but the d-proline in all three is the enantiomorph of the I-proline obtained by protein hydrolysis. Butyryl formic (=dimethyl pyruvic) and pyruvic acids are * Translation. NoTE.-On Corn Spurs Latin Clavi siliginis (Nails of Com) :-There are often found on the ears of rye or corn long black hard narrow spurs being close to and between the corn that is to say in the ears growing out of them and extending out so that they look like long nails; they are white inside like corn and are quite harmless to the corn.Such corn spurs are regarded by women as a particularly useful and valuable remedy for swelling and pains of the womb; for this purpose they are taken about three times and eaten BARGER THE ALKALOIDS OF ERGOT 343 further closely related to the amino acids valine and alanine respectively and lysergic acid is doubtless derived from tryptophane by the addition of a chain of five carbon atoms which may represent yet another amino acid residue. The hydroxyisopropylamine of ergometrine may be regarded as a reduction product of alanine ("alanol" ) . It is necessary now to consider the constitution of lysergic acid the most characteristic fission product and the only one common to all four pairs of alkaloids.Lysergic acid and its simple amide ergine have been obtained in two forms corre-sponding to the two series of potent and of inert alkaloids. The following table illustrates this by the values of [a]546.1 in pyridine according to Smith and Timmisn: Ergometrine - 16" Ergometrinine + 596" Iso-ergine + 25" Ergine + 635" Lysergic acid +49' Iso-lysergic acid + 365" It will be seen that the ergine first obtained belongs to the ergotinine series, and the (first) lysergic acid originally discovered by Jacobs and Craig is analogous to ergotoxine. Ergometrine and ergometrinine can be partially converted into each other by boiling N alcoholic potassium hydroxide solution ; some hydrolysis to lysergic acid always occurs but is evidently slower than the conversion into the other alkaloid (Smith and Timrnis'O).The isomerism of these substances and of the other pairs of ergot alkaloids clearly depends on the lysergic acid portion and seems to be determined by a shift of a double bond the only one known to occur in this acid. Jacobs and Craigsg reduced (laevorotatory) ergotoxine ergotamine and ergometrine catalytically and then obtained a-dihydrolysergic acid [a] - 106" also obtained by reducing lysergic acid itself; on the other hand the dextrorotatory ergotinine and ergotaminine are after reduction hydrolysed to y-lysergic acid [a] +33". (6-Lysergic acid is yet another substance obtained in a different way.) Jacobs and Craig consider that in the dihydrolysergic acids a stereoisomerism has become fixed by the abolition of the double bond; the dihydro acids are not interconvertible; the methyl ester of lysergic acid shows mutarotation in hot methyl alcohol the esters of the dihydro acids do not.The constitution of lysergic acid has not yet been fully established but a formula suggested by Jacobs and CraigSQ is doubtless so near the truth that it may serve to illustrate the main features of the structure. 8 7 H&-CH. COOH / \ \ / 9 H2C D N.CH 6 10 c=c 5 Rings A and B with carbon atoms 4 and 5 and the upper nitrogen form a trypt-amine residue which has been isolated as such from a number of other alkaloids (calycanthine evodiamine strychnine).The a-position of the pyrrole ring (2 344 BARGER THE ALKALOIDS OF ERGOT is unsubstituted which accounts for the Hopkins and Cole and other colour reactions of the ergot alkaloids. The much employed reaction with +-dimethyl-aminobenzaldehyde is however so intense and peculiar that it may well be conditioned by a feature which has so far not been encountered in alkaloidal chemistry namely the ring closure with hydrogenated C, of the benzene ring; in all other indole alkaloids of known constitution this carbon atom is not shared by a second ring. There are two reduced rings C and D which on distillation with soda lime survive together as quinoline. Potash fusion of dihydrolysergic acid yields 1-methyl-5-amino naphthalene in which rings A and C survive with carbon atom 9 (the methyl group) and the indole nitrogen atom.Oxidation of lysergic acid yields a tribasic acid CI4HgO,N in which carbon atoms 2 12 and 15 become carboxyl-ated and the original carboxyl at 7 becomes combined with the nitrogen at 6 to form a quinoline-betaine. The latter nitrogen bears the one N-methyl group of ergot alkaloids and is the only basic one; the carboxyl group at 7 is present in an amide grouping in the alkaloids. The position of the double bond is placed tentatively between C and Cl0 and Jacobs and Craig imagine that in mutarotation it shifts to 9-10 and/or 4-5 but its position in the two lysergic acids and their relationship is still uncertain. The rapid and remarkable advance in our knowledge of the ergot alkaloids during recent years has as it were rounded off the subject and has allowed of a brief presentation of the main results in tabular and diagrammatic form.It may perhaps be of interest to consider now also the difficulties which retarded progress in the past. The possibility of isomerism in the lysergic acids not only gives rise to opposite optical rotation and to a great difference in pharmacological potency in the two series of alkaloids but also brings about a difference in their power of crystallising. The inert alkaloids are sparingly soluble in organic solvents and crystallise readily without solvent of crystallisation. The potent alkaloids, on the other hand have a remarkable power of forming additive “compounds” with the solvent and with isomeric inert alkaloids (ergoclavine and sensibamine are such crystalline molecular “compounds”).It is therefore natural that the first alkaloid ergotinine crystallised by the French pharmacist Charles Tanret in 1875,a belongs to the almost inert series and that he regarded a potent alkaloid present in the mother liquors of ergotinine as merely an amorphous and impure variety of the crystalline. This explains a prolonged controversy about the pharmacological action of ergotinine which controversy becomes the more intelligible when we consider that ergotinine is readily transformed into the potent ergotoxine. The high molecular weight and residual valency of the latter alkaloid for a long time prevented its recognition as a separate substance and during the three decades which followed Tanret’s discovery pharmacologists (Kobert,46 Jacobj29) attributed the gangrenous and other actions of ergot not to an alkaloid, but to amorphous acidic and phenolic substances which as we now know owed their activity to adsorbed ergotoxine; the latter is a weak base having also acidic properties.Dale,lo indeed began his fundamental work on the pharmacology of ergot with such a substance (chrysotoxin) containing about 2 per cent. of adsorbed ergotoxine. The separation of the latter is complicated by the fact that the salts of ergotoxine are little soluble in water and are precipitated from their colloida BARGER THE ALKALOIDS OF ERGOT 345 solutions by electrolytes notably by excess of mineral acids. Strong acids should accordingly be avoided in the chemical assay of the alkaloids which have often been weighed mixed with non-alkaloidal substances.Ergotoxine the most abundant active alkaloid in ergot was isolated as crystalline phosphate by Barger and Carr.495 Its existence was recognised independently by Kraft,46 who obtained from ergotinine the sparingly soluble sulphate of another alkaloid which he regarded as formed from ergotinine by hydration (hydroergotinine). The additional molecule of water seemed to be confirmed by analyses of crystalline ergotoxine salts (Barger and Carr5) but I now incline to the view of Stoll and others that ergotoxine is not a hydrate of ergotinine but isomeric with it. Much later ergotoxine was crystal-lised as free base by Smith and Timmisu; for this purpose it should be liberated from its salts by sodium bicarbonate or borate not by sodium carbonate and the only suitable solvents for crystallisation are benzene and carbdn bisulphide.These are so tenaciously retained in the crystals that even now no concordant analyses of this alkaloid have been published. Ergotamine (StolP) crystallises with a variety of solvents-particularly well from aqueous acetone with 2H,0.2C3H,O. Its salts are generally rather more soluble than those of ergotoxine. In alcoholic solution it shows mutarotation (Frhrejacque and Raymond-Hametl7) with partial conversion to ergot aminine (StolP). The latter alkaloid crystallises without solvent in characteristic triangular leaflets and is probably the least soluble of all ergot alkaloids (1 6400 in alcohol at room temperature).Ergosine and ergosinine were discovered by Smith and Timmiss9; I am greatly indebted to these authors for further details from a paper since published''. Ergometrine differs from the more complex alkaloids in being moderately soluble in water and only very sparingly soluble in chloroform. This and its lack of characteristic pharmacological properties account for its having been overlooked for so long. Its existence in aqueous extracts of ergot was first recognised through the clincial experiments of and it was then isolated by (Dudley and) Moir.14 As the result of Moir's first publication a water-soluble principle was sought for and found by several other workers; the crystalline substance was described in suc-cessive months (March-June 1935) as ergometrine by Dudley and Moir as ergo-tocine by Kharasch and Legault,G as ergobasine by Stoll and Burckhardt,sl and as ergostetrine by Thompson.*5 After an exchange of specimens these various substances have been recognised by those concerned to be identical (Kharasch, King Stoll and Thompson44).There seems little doubt that the priority both biological (1932) and chemical (1935) belongs to Dudley and Moir and that the alkaloid should therefore be called ergometrine. It crystallises with a variety of solvents some of which (benzene methyl ethyl ketone) are given off completely in a high vacuum at 100" C. and when thus freed from solvent ergometrine melts at 162-3" C. It also crystallises in this condition from ethyl acetate at -4" C., but at room temperature it forms different crystals with 0-5 mol.of ethyl acetate, which is so tenaciously held that this second variety melts with effervescence, but without darkening at 130-132°C. This ethyl acetate compound is more stable to atmospheric oxygen than the crystals containing benzene or ethyl methyl ketone (Dudley13). By the action of acids or alkalis ergometrine is partiall 346 BARGER THE ALKALOIDS OF ERGOT converted into the inert ergometrinine (Smith and T i m m i ~ ~ ~ ) which is much less soluble in water and much more soluble in chloroform than ergometrine into which it can be partially re-converted by acids and alkalis. It crystallises without solvent of crystallisation. The remarkable difference in pharmacological properties optical rotation and power of associating with solvents which the two series of ergot alkaloids exhibit as a result of an isomeric change in the lysergic acid portion of the molcule is a phenomenon of considerable interest and excites curiosity as to what exactly is the nature of the isomerism.The residual valency which the active series evidently possesses is perhaps also responsible for the pharmacological action by anchoring the alkaloids to the cells on which they act. Further chemical knowledge of the constitution of lysergic acid may perhaps ultimately lead to a deeper insight into the reason why drugs act. Ergoclavine (Kiissne+) is probably an equimolecular mixture of ergosine and ergosinine and sensibamine (Chinoin A.G. and Wolfg) a similar mixture of ergotamine and ergotaminine. An alkaloid C,,H,,O,N is described by Holden and Driver,27 who state that it is not an indole derivative.The first third and fourth pairs of alkaloids seem to occur in all specimens of ergot of rye but it should be emphasised that the second pair is absent from most specimens of the drug and occurs only in ergot of rye from a particular source. Their discoverer Stoll has declared this source to be “Central Europe,” and we may take it to be Hungary for sensibamine (according to Stoll a mixture of ergotamine and ergotaminine) was discovered there and Hungarian ergot was the only one reported on by Kiissnefl8 which yielded these alkaloids. Smith and Timmis examined a large number of commercial specimens of ergot of rye from many countries and always isolated the first pair of alkaloids and never the second.They obtained,a however ergotamine and ergotaminine and these only from a non-official ergot Claviceps fiuurpuurea growing on Festuca elatior in New Zealand. They employed the simple ether extraction method of Kraft, and accordingly consider that the isolation of ergotamine and ergotaminine does not depend on the special method of extraction devised by Stoll but on the nature of the ergot. It would be interesting to know more about the biological peculiarities of the Hungarian ergot yielding these alkaloids. Since the two alkaloids in the various pairs are readily transformed into each other it is a matter of doubt to what extent each occurs as such or arises during the process of separation. Stoll considers that ergotaminine does not occur in ergot as such and is a secondary product formed during extraction.Ergometrinine has only been obtained with difficulty from ergometrine mother liquors and may be formed in similar fashion. The relative proportion of the other alkaloids is by no means constant. Kussner,48 before the discovery of ergometrine gives the following examples as percentages of total crude alkaloid actually crystallised : Ergotoxine Ergotinine phosphate Ergoclavine Total Spanish . . 37.6 11.5 19.3 68.4 Do. 31.2 15.5 18.5 65.2 Do. 29.0 18.7 15.8 63.2 Russian . . 4595 10.0 19.9 7 5 BARGER THE ALKALOIDS OF ERGOT 347 On the assumption that ergoclavine is an equimolecular mixture of ergosine and ergosinine the ratio of inactive to active alkaloid varies from 1.4 to 2.2, but it may well be lower.The amount of ergometrine in Spanish ergot is con-sidered by Smith and Timmis (private communication) to be 10 to 15 per cent. of the total but assays by Hampshire and Page24 indicate a somewhat higher pxopor-tion (about 20 per cent.). The pharmacologically potent alkaloids (ergotoxine ergotamine ergoclavine, ergosine) all antagonise the action of adrenaline in much the same way and since chemical and physical assay methods cannot distinguish between these and their almost inert isomerides the only reliable way of standardising ergot preparations is by biological methods best by that of Broom and Clark’ on the isolated rabbit’s uterus. The chemical determination of total alkaloid indeed gives a useful rough indication of the value of a particular ergot for the proportion of inactive to active alkaloid does not vary excessively (see above).The discovery of ergometrine the therapeutically important principle has further complicated the question for there is no pharmacological method for its assay. Again on the assumption that the proportion of ergometrine to the other alkaloids does not vary greatly (it seems to be something like 1 6) chemical assays of total alkaloid give some indication but an accurate chemical method of assaying ergometrine separately is very desirable and seems quite possible (see below). CHEMICAL AND PHYSICAL METHODS OF ASSAY.-These may be divided into: (1) Chemical methods in which the alkaloids of the first three groups are either weighed or titrated. (2) A colorimetric method depending on the lysergic acid grouping which, therefore seems to give identical colorations for equimolecular amounts of all four groups (including ergometrine).(3) Spectrophotometric methods which do the same as (2). (4) A method assaying separately the water-soluble alkaloid ergometrine. (1) ChemicaE Methods.-The oldest is that of the Swiss pharmacist Kelle~-:l?~~ and the prototype of most present-day processes of alkaloidal assay depending on the extraction from ethereal solution by acid and after making alkaline shaking out again with ether evaporating and weighing the residue. This process as applied to ergot has been modified by Fromme,20 but even after this modification it is open to objection. Meulenh~ff~s~ and Harm~ma~59~~ have pointed out that when the first ethereal solution is shaken with hydrochloric acid a crystalline precipitate of an alkaloidal hydrochloride is formed which Fromme filters off and discards; hence tartaric acid should be used instead of mineral acids.(Ergotoxine salts are precipitated by electrolytes.) A much more serious error in the opposite direction has been pointed out by Oette15’ and by van Pinxtereii5g the final ethereal residue may contain much inert matter and the results may be much too high. This error affected the older assays which sometimes indicated an alka-loidal content as high as 0-4 per cent. never found in recent years. It is difficult to separate the alkaloids from a yellow colouring matter (see above JacobjZ9 and Kraft46). If the alkaloids are precipitated by sodium carbonate and filtered off 348 BARGER THE ALKALOIDS OF ERGOT the results are lower and near the truth; this procedure is adopted by the German Pharmacopoeia in order to eliminate water-soluble amines (Gadamer and Neuhoff21).Whilst the amines do not appreciably pass into ether the filtration favours the separation of alkaloids from the yellow colouring matter which is soluble both in ether and in sodium carbonate solution. The use of ammonia has been suggested for the liberation of the alkaloids but since ergotoxine begins to dissolve at pH 8, this should be avoided; a buffer mixture (0.2 mol. of disodium hydrogen phosphate + 0.1 mol. of citric acid) is recommended by van Pinxteren. The latter author prefers weighing the alkaloids to the titration recommended by the German Pharmacopoeia on the ground that the voluminous precipitate adsorbs sodium carbonate which owing to its equivalent being one quarter of that of the alkaloids, causes when included in the titration an error four times as great as that due to weighing it.Schlemmer Wirth and Peters62 recommend a micro-Kjeldahl estimation of the crude alkaloidal residue; here all five nitrogen atoms of the alkaloids are converted into ammonia and give five times the titration value of the alkaloids themselves which are monacidic bases ; moreover the non-nitrogenous colouring matters do not affect the result after destruction by the Kjeldahl method. With the above-mentioned modifications the Keller-Fromme method like that of the German Pharmacopoeia is satisfactory in the opinion of Van Pinxteren and of Schlemmer Wirth and Peters ; compare also Leinziger and Kelemen,60 Prybill and Maurer,sO as well as Gatty-Kostyal and Derlatka.22 The procedure of the German Pharmacopoeia has been chiefly criticised by OettelJ5' who has adapted for quantitative use a somewhat peculiar method of isolating the alkaloids (Forst15).With 50 per cent. aqueous alcohol or acetone the alkaloids are extracted from ergot (which is not de-fatted) ; as much as 500 grams is used. The evaporation of the alcohol or acetone even at a low temperature leads according to van Pinxteren to decomposition of alkaloids and to low results. Schlemmer Wirth and Peters to some extent confirm this but do not consider the error to be serious. In any case the Oettel process is complicated.In the method of Goris and Liot23 the alkaloids are precipitated with silicotung-stic acid; the precipitate is ashed and the weight of ash multiplied by 0.74 gives the weight of alkaloid. According to van Pinxteren this method has not received the attention which it deserves. In dilute solution the reagent causes only a colloidal turbidity (Harmsma) but according to van Pinxteren the suspension coagulates in an ice-chest and can then be collected in a Gooch crucible washed and ignited. (2) Of late a colorimetric method has come into vogue and has been adopted by the British Pharmacopoeia. The original colour reaction of Tanret greatly improved by Keller,42 was made still more sensitive and specific by the use of 9-dimethylamino-benzaldehyde (van Urks7).It thus becomes a true indole reaction and can be used quantitatively (M. I. Smith@). The blue colour is due to the indole complex in lysergic acid and the intensity of the colour is the same for equimolecular amounts of all the ergot alkaloids. In the original method the blue colour is only developed fully on exposure to light or heat; hence Allport and Cocking1 have suggested the addition of a trace of ferric chloride (first used in a different way by Keller) which at once gives the maximum colour. This modification is preferred e.g. by Upsher Smithm and others. Freudweilerls 91 BARGER THE ALKALOIDS OF ERGOT 349 uses vanillin instead of p-dimethylamino-benzaldehyde ; the red colour produced by the former aldehyde is said to be more intense than the blue due to the latter, and exposure to light is unnecessary for its development.For other papers on the colorimetry of ergot alkaloids see Lozinski Holden and Diver,51 Swoap Cartland and Hart,m Stevens,76 and Swanson Powell Stevens and StuartE2; these authors are all concerned with a comparison between the colorimetric and a biological method. (3) The ultra-violet absorption spectrum of ergotinine was first determined by Brustier,8 and used for the assay of the alkaloids by Harrn~rna,2~9~~ van Itallie,28 Wokes and Crocker,ss Allport and Crews.2 There is a maximum at 316-318mp and a minimum at 272mp; practically the same curve has been found for all the alkaloids (for ergometrine by Allport and Crews; ergosinine has not been investi-gated). The advantage of the spectroscopic method is its delicacy (even 1 part in 500,000 can be estimated) ; the solutions employed need be only moderately pure.A comparison of the various methods so far referred to has been made by van Pinxteren and by Schlemmer Wirth and Peters. (4) The Separate Assay of Ergometrine.-This has so far only been attempted by Hampshire and Page,24 and apparently with some degree of success. The three methods of assay above referred to do not distinguish between ergometrine and the more complex alkaloids. Hence the rather pronounced differences in solubility between the former and the latter must be utilised. Hampshire and Page found that certain solvents (carbon tetrachloride and amyl ether) extract ergotoxine ergotamine etc. from pure aqueous solution but not ergometrine ; when applied to the impure mixture of alkaloids directly extracted from ergot the separation was however incomplete.On the other hand it was found possible to remove ergometrine and this alkaloid only from a crude mixture of total ergot alkaloids in ether by repeated shaking with water until the aqueous layer no longer gave the colour reaction with dimethylamino-benzaldehyde. The combined aqueous extracts and the alkaloids remaining in the ether are then separately estimated colorimetrically against ergot oxine as standard and the amount of ergometrine is obtained by multiplying by the factor 0.538 (ratio of molecular weights). As an example 10 g. of ergot yielded 0.0128 g. total alkaloids, of which 0.0019 g. was water-soluble. Added ergometrine was recovered satis-factorily.The percentage of ergometrine in the total alkaloidal content was found for samples of Spanish ergot to be 24 16 20 20 and 19 per cent. for Russian 13 and 21 per cent. ToxIcoLocY.-Although in larger doses ergometrine is poisonous at least to animals the characteristic symptoms of ergot poisoning must be mainly attributed to the more complicated alkaloids ergotoxine ergotamine and ergosine which by constriction of the blood vessels impede the circulation and thus cause gangrene. This condition has been produced experimentally in the cock's comb by repeated intramuscular injections of pure ergotoxine and ergotamine. This production of gangrene by pure alkaloids has been attempted only in a few other species and mostly with little success.The alkaloids also have certain effects on the central nervous system which lend further support to the view that they were the active agents in the epidemicsof human ergotism which raged in the Middle Ages and hav 350 BARGER THE ALKALOIDS OF ERGOT Fig. 2 Woodcut of St. Anthony with a victim of gangrenous ergotism, who has lost his right foot; the left hand burns with symbolical flames of St. Anthony’s fire. From Gerrsdorff’s book on military surgery 1535. * Translation. ON “THE COLD FIRE” (GANGRENE) w h y dost thou claim the power to heal? No longer now to any man. 0 Anthony thou holy man, To God the Lord the honour’s due occurred in a much milder form and on a much smaller scale even in recent years. Our knowledge concerning the toxic substances responsible for ergotism is from the very nature of the disease insecurely founded and we are reduced to hypotheses and speculation.There were two types of the disease which in their initial stages had a number of milder symptoms in common. The gangrenous type was characterised by violent burning pains in the part affected (more often a foot than a hand); hence the terms “fire of St. Anthony” and “holy fire,” applied to the disease. After a time tbe pain was apt to stop the diseased part became numbed and cold and often black (“like charcoal,” as the chronicles have it) and finally the mummified limb was generally detached at a joint without pain or loss of blood. The extent of the gangrene varied from the mere shedding of nails and the loss of fingers or toes to that of several entire limbs.Apart from this loss there were no other serious symptoms and even in moderately severe cases the patient made a good recovery when placed on a wholesome diet. Fig. 3 Fig. 4 Frescoes in the Chapel of St. Anthony Mural paintings from the chapel of St. Anthony at Waltalingen near Zurich (second half of fifteenth century). Fig. 3 refers to gangrenous ergotism (cf. Fig. 2). Fig. 4 represents an appeal to the Saint by sufferers from convulsive ergotism (cf. especially the hands with Figs. 6 and 7). Convulsive ergotism was on the other hand not a vascular but a nervous disease characterised by twitching and epileptiform convulsions in the severest cases by delirium blindness deafness paralysis and insanity. After one of the last epidemics (in Germany 1879) a degeneration of the spinal cord was demon-strated by modern histological methods.It is a remarkable fact that although sometimes (e.g. in Russia) mixed epidemics occurred gangrenous ergotism was confined chiefly to France and the convulsive variety to Germany. An explanation may be sought tentatively in a difference in the diet of the poorest peasants of the districts affected. Accordin 352 BARGER THE ALKALOIDS OF ERGOT to this hypothesis gangrene is the primary effect of a prolonged administration of massive doses of the ergot alkaloids which impede the circulation; the arterioles, as modern pharmacology has shown become constricted. In convulsive ergotism a second factor would intervene and produce nervous symptoms with doses of ergot not in themselves capable of causing gangrene.This second factor is most probably a deficiency of the fat-soluble vitamin A in the diet E. Mellanby62 has shown that ergot and to a lesser extent wheat germ produces lesions in the spinal cord of dogs symptoms somewhat analogous to nervous ergotism in man. Probably the active agent of the ergot in this case is not an alkaloid but a simpler substance. If the above hypothesis as to the cause of convulsive ergotism is correct it has to be shown that the diet of the poor French peasant was richer in vitamin A than that of the German. The literature of the eighteenth century on epidemics in these two countries supplies proof that such was the case. Even in the Sologne, formerly a damp and arid district of France (south of Orleans) where gangrenous ergotism was endemic there were enough pasture cows milk and butter.In the North German plain ( e g . in Hanover) contemporary records show that dairy produce was lacking in the diet of the labourers and certain details in the descrip-tion of German epidemics have almost the evidential value of a control experiment in showing that a shortage of vitamin A was a contributory cause of the disease. For details concerning this point and for a full historical account of ergotism, based as far as possible on contemporary sources I would refer to my monograph on “Ergot and Ergotism”S from which the figures are reproduced by kind permission of the publishers. Certain problems still remain unsolved and it is not even possible to advance hypotheses concerning them.Thus whilst it is certain that rye was largely grown in England in the Middle Ages ergot was so rare that there is no true English name for the fungus. (In the above monograph I have quoted close on 50 German names and some 17 French names.) Another question is why there are unmistakable references to gangrenous ergotism (in French chronicles) from the ninth and tenth century onwards and yet there is no clear indication or description of convulsive ergotism in Germany before the end of the sixteenth century. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. N. L. Allport and T. T. Cocking Quart. J . Pharm. 1332 5 341. N. L. Allport and S. K. Crews id. 1935 8 447. G. Barger Ergot and Ergotism 1931 (London Gurney & Jackson).G. Barger and F. H. Carr Chem. News 1906,94 89. - J . Chem. SOC. 1907 91 337. G. Barger and A. J. Ewins id. 1910 97 284. W. A. Broom and A. J. Clark J . Pharm. exp. They. 1923,22 59. V. Brustier Bull. SOC. Chim. 1926 [iv] 39 1538. Chinoin A.G. and Wolf Upjest (Anmeld. C. 45550 K1. 12 p. 11). H. H. Dale J . Physiol. 1906 34 163. H. W. Dudley Brit. Med. J. 1935 i 798. - Pharm. J.; 1935 80 709. - Pvoc. Roy. Soc. London 1935 1 1 8 ~ 478. H. W. Dudley and J. Chassar Moir Brit. Med. J. 1935 i 520. A. W. Forst Arch. exp. Path. Pharm. 1926 114 125. M. Frdrejacque and Raymond-Hamet Rev. d . Pharm. th&. exp. 1!)29 1 333. - id. 1936 181 180 Fig. 5 Gangrene of the Toes due to Ergot. Hungary 1908 Fig. 6 Contractures of Hands and Feet.From a dissertation by Heusinger describing the epidemic near Marburg 1856 1 Fig. 7 Fatal Case of Convulsive Ergotism. Hungary 190 BARGER THE ALKALOIDS OF ERGOT 353 18. R. Freudweiler L'Ergot de Seigle ses Principes actives et leurs Dosages Inaug. Diss., Zurich 1932. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. * 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. - Pharm. Acta Helvet. 1932 7 116 139. G. Fromme Jahresber. Caesar und Loretz A.G. 1915. J. Gadamer and E. Neuhoff Arch.Pharm. 1926 264 546. M. Gatty-Kostyal and P. Derlatka Bull. Sci. Pharm. 1929 37 471 (Abstr.). A. Goris and A. Liot Bull. Sci. Pharm. 1924 31 379. C. H. Hampshire and G. R. Page Quart. J . Pharm. 1936,9 60. A. Harmsma Kwantitatieve Bepaling van het absorbeerend Vermogen van de Moederkoorn-alkaloaden in het Ultraviolette Gebied en een practische Toepassing daarvan Diss. Leiden 1928. - Pharm. Weekbl. 1928 65 1114. G. W. Holden and G. R. Driver Quart. J . Pharm. 1936,9 231. L. van Itallie Schweiz. Apoth.-Ztg. 1928 66 423. C. Jacobj Arch. exp. Path. Pharm. 1897 33 85. W. A. Jacobs and L. C. Craig J . Biol. Chem. 1932 97 739. - id. 1934 106 393. - id. 1935 110 521, - Science 1935 82 16. - J . Amer. Chem. Soc. 1935 57 960. - J. Biol. Chem. 1936 113 767.- ad. 1936 115 227. - J . Org. Chem. 1936 1 245. C. C. Keller Schweiz. Wochenschr. f. Chem. u. Pharm. 1894,32 121 133. - I uber die Wertbestimmung von Drogen und galenischen Praparaten; 11 Neuere Studien iiber die Bestandteile des Secale cornutum Inaug. Diss. Zurich 1897. M. S. Kharasch and R. R. Legault Science 1935 81 338. M. S. Kharasch H. King A. Stoll and M. R. Thompson Nature 1936 137 403. R. Kobert Arch. ex$. Path. Pharm. 1884 18 316. F. Kraft Arch. Pharm. 1906 244 336. W. Kussner E . Merck's Jahresbericht for 1933 1934 47 5 (D.R.P. 606,778). - Z. angew. Chem. 1937 50 34. M. von Leinziger and J. von Kelemen Arch. exp. Path. Pharm. 1928 128 173. E. Lozinski G. W. Holden and G. R. Driver Quart. J . Pharm. 1933 6 395. E. Mellanby Brit. Med. J. 1930 i 677.J. S. Meulenhoff Bey. NederZ. Maatsch. t. bevord. d. Pharm. Achtste Volgreeks No. 1, The Hague 1899. - Pharm. Weekbl. 1902 89 101. J. Chassar Moir Brit. Med. J. 1932 i 1119. - id. 1936 ii 799. H. Oettel Arch. ex?. Path. Pharm. 1930 149 218. J. A. C. van Pinxteren Phavm. Weekbl. 1931 1151. - id. 1934 1230. A. Prybill and K. Maurer Arch. Pharm. 1928 266 464. F. Schlemmer and H. Schmitt Arch. Pharm. 1932 270 15 29. F. Schlemmer P. H. A. Wirth and H. Peters id. 1936 274 16. M. I. Smith U.S. Public Health Reports Washington 1930 1466. S. Smith and G. M. Timmis J . Chem. SOL 1930 1390. - id. 1931 1888. - id. 1934 104 547. - id. 1935 108 594. - id. 1935 111 455. - id. 1907 245 644. - id. 1932 763. - id. 1932 1543. - id. 1934 674. - Nature 1936 137 1075. - J. Chem. Soc. 1936 1166. - id. 1936 1440. - id. 1937 396. F. A. Upsher Smith J . Amer. Pharm. Assoc. 1934 23 25. A. Soltys Ber. 1932 6 5 ~ 553. F. Sternon and Rensonnet Compt. rend. XIIe Cong. intern. Pharm. 1935 237; from A. N. Stevens J . Amer. Pharm. Assoc. 1933 22 940. A. Stoll Verhandl. d. Schweiz. Naturf. Ges. 1920 190. - Schweiz med. Wochenschr. 1935 ii 1077. - Wiener klin. Wochenschr. 1936,49 1513 1552. A. Stoll and E. Burckhardt Conzpt. rend. 1935 200 1680. Quart. J . Pharm. 1936 9 307 354 MORGAN AND PRITCHARD THE AVERAGE VITAMIN A 81. 82, 83. 84. 85. 86, 87. 88. A. Stoll and E. Burckhardt Bull. Sci. pharnaac. 1936 42 257. E. E. Swanson C. E. Powell A. N. Stevens and E. H. Stuart J . Amer. Phnrm. Assoc., D. F. Swoap. G. F. Cartland and M. C. Hart id. 1933 22 8. Ch. Tanret Compt. rend. 1875 81 896. M. R. Thompson Science 1935 81 636. E. E. Turner Ann. Reports Chem. Soc. 1935 32 35. H. W. van Urk Pharm. Weekbt. 1929 66 473. F. Wokes and H. Crocker Quart. J . Pharm. 1931 4 420. 1932 21 229 320 1003

ISSN:0003-2654    

DOI:10.1039/AN9376200340

出版商:RSC    

年代:1937

数据来源: RSC

摘要:

OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc.it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned.In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix.about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary.Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary.It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order.It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained.In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents. To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'.It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air.There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary.It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined.It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'.It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place.The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C.It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined.It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'.It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place.The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix.about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary.Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air.There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned.In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order.It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary.Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary.It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order.It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained.In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air.There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary.It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary.Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,

ISSN:0003-2654    

DOI:10.1039/AN9376200354

出版商:RSC    

年代:1937

数据来源: RSC

摘要:

EVANS THE DETERMINATION OF COBALT 363 The Determination of Cobalt. A new Method of Volumetric Determination and a new Method for its Determination in Steel* BY B. S. EVANS M.B.E. M.C. DSc. F.I.C. GRAVIMETRIC.-The literature of the analytical chemistry of cobalt is ex-tensive complicated and many-sided. Few elements have more or more multifarious methods of determination to their credit; on the other hand few have so many methods which are not quite good enough when one requires great accuracy. The electrolytic method beloved of text-books seems either not to remove all cobalt from solution or to give a contaminated deposit? The phosphate precipitate obtained by Schoeller and Powell's method2 leaves a certain amount of cobalt in solution for which a correction has to be made; precipitation as sulphide appears to be complete but the problem of washing the precipitate is acute; water carries it through the filter ammonium salt solutions dissolve appreciable amounts, alkali salt solutions are satisfactory but fatal to a gravimetric finish.The precipitation with a-nitroso-p-naphthoP5 appears to be complete and the pre-cipitate can be washed; also it has the merit of giving a separation from nickel; the difficulty here lies in the fact that it is (or was) apparently impossible to buy a reagent uncontaminated with iron. The finish of the determination is a further slight difficulty; doubts have been thrown (somewhat unnecessarily I consider) on the accuracy of the conversion to Co,O on ignition.814 On the other hand one may convert the cobalt into CoSO and weigh it as such but here again there is the doubt whether it is possible to eliminate all excess sulphuric acid without at the same time decomposing a trace of the sulphate495.In my view however with careful work both oxide and sulphate methods are capable of greater accuracy than has generally been assigned to them. Reduction to metal by ignition in a current of hydrogen is said to overcome the difficulty in both cases.39 A method has also been published for obtaining the cobalt complex with a-nitroso-#3-naphthol free from the precipitant and weighing it as such.34 Good results both gravimetric and volumetric appear to have been obtained by the 8-hydroxyquinoline So many metals are precipitated by this reagent however that a rigorous separation of the cobalt is first required.SEPARATION.-The separation of cobalt is in itself not an easy matter; it co-precipitates with stannic tin when the latter is thrown down by hydrogen sulphide in hydrochloric acid solution,g and with zinc when the precipitation is made in acetic acid ~olution.~ In view of these facts and of the work published showing that iron and zinc both co-precipitate with copper sulphide in quite strongly acid solution,1° there is at least a suspicion in lack of positive evidence to the con-trary that co-precipitation of cobalt with sulphides is a thing to be guarded against. Cobalt co-precipitates to a serious extent with metals of the third group thrown down with ammoniall. It of course follows nickel in most of its reactions. If cyanide is used at any stage for keeping cobalt from being co-precipitated in some reaction there is danger that some at least of the cobalt will be converted into * Communication from the Research Department Woolwich 364 EVANS THE DETERMINATION OF COBALT cobalticyanide in which form not only is the cobalt difficult to recover but the cobalticyanide itself gives precipitates with many metals in acid solution.One of the most difficult and also one of the most important metals to separate from cobalt is iron and here the valuable a-nitroso-/3-naphthol process as hitherto applied breaks down for the reagent precipitates the iron as well as the cobaltJ2 In consequence of this the usual procedure is to separate the cobalt from the iron group by precipitating the latter with zinc oxide assuming that the cobalt is left entirely in solution.This point has been investigated by Hoffman,ls who came to the conclusion that although the method gives approximate separation a certain amount of cobalt is always precipitated and he accordingly gave a method for eliminating this error but apart from the somewhat irksome restrictions in carrying out the precipitation the method demands re-precipitation which in steel analysis involves the solution and retreatment of the ferric hydroxide from 2 grams of steel. It is noteworthy that Hoffman followed up his investigation into the zinc oxide separation by a second paper a year later giving a method for determining cobalt in steel in which this separation is not used a t all the iron being removed by ether extractionJ4 A useful method for the separation of cobalt by means of precipitation with phenyl-thiohydrazoic acid eliminates many metals including the bulk of the iron, but not apparently nickel or the last few mg.of iron.4 The dinitroso-resorcinol precipitation requires the removal of iron and copperJ6 VOLUMETRIC.-Turning next to volumetric processes for the find determination of the separated cobalt we have the choice of a large number of methods. Of these, some depending on precipitation by 8-hydroxyquinoline etc. appear as mentioned above to be capable of giving good results. The same seems to be true of certain electrometric methods16; these latter however need of course special apparatus and technique. The attention of those in search of volumetric processes for cobalt has been largely centred on the fact that precipitated cobaltic hydroxide can be dissolved in a measured excess of a reducing agent and the excess of the latter titrated back.A number of methods of applying this reaction were examined by Willard and Hall who published many figures obtained by various modifications; many of these figures are excellent but on the whole the variations seem to fall well outside those obtained by the method about to be described. The process has some distinct drawbacks chiefly due to cobaltic hydroxide being a substance which does not dissolve readily in ordinary reducing agents. Willard and Hall's conclusions were that with potassium iodide the reaction was very slow with acid ferrous sulphate an empirical factor had to be used and with stannous or titanous chlorides air must be rigorously excluded.One of the most recent methods dependent on this principle is that of Sarver?' who reduces with ferrous sulphate in the alkaline liquid air of course being excluded. I have tried many experiments with a view to making the cobaltic hydroxide method readily workable but without obtaining consist en t 1 y accurate results. Other volumetric methods proposed are those based on titration of precipi-tated cobaltinitrite,l*J9Jo cobalt pyridine t h i o ~ y a n a t e 2 ~ ~ ~ ~ ~ ~ ~ or cobalt x a l a t e ~ ~ . All these processes like that using hydroxyquinoline involve filtration as a EVANS THE DETERMINATION OF COBALT 365 integral part of the volumetric determination and some need corrections for solubility.In another recent method cobalticyanide is titrated with silver nitrate, potassium chromate being used as an i n d i ~ a t o r . ~ ~ In face of all this wealth of processes it may seem waste of time putting forward yet another; there is however one possible method which has for reasons to be stated later been very much neglected of late years-the cyanide titration. THE CYANIDE TITRATION.-oVer and above the fact that most if not all, reliable methods of volumetric determination of cobalt involve filtI ation and its attendant evils the cyanide method has stood in need of investigation because it is perhaps the best method of determining nickel and hence the analyst has been faced with a troublesome separation of nickel from cobalt when as frequently happens the two occur together.One outstanding paper has been published of recent years-that of Glasstone and Speakman26; the authors investigated the cyanide titration electrometrically and confirmed the theoretical basis of Rupp and Pfenning’s method,27 which assumes a Co (CN) ratio of 1 5 at the end-point. Rupp and Pfenning’s method involves running the solution of nickel and cobalt into the cyanide-a procedure which though quite simple and straightforward when it is essential is one that involves an additional dilution to a known volume, with a certain amount of waste of a possibly limited sample and one instinctively avoids it where possible; in addition the end-point obtained is apparently difficult to observe and conditions as to alkalinity amount of ammonium salts etc.have to be somewhat carefully adjusted. The Rupp and Pfenning titration of course gives the sum of the nickel and cobalt and to obtain a measure of the individual com-ponents Glasstone and Speakman were driven back on the cobaltic hydroxide method which has been criticised above. A full discussion of the theoretical aspect of the question is given in the paper referred to and there is no point in repeating it here. The authors however dismiss their attempt to use the ordinary potassium cyanide-silver iodide titration for cobalt with the statement that they found it impossible to get an end-point. What .actually happens is that when one has titrated the liquid with potassium cyanide until it is clear after a moment’s stand it clouds again; the liquid can again be cleared with cyanide but it again clouds this time more slowly and so on, the recurrence of the clouding becoming more and more slow ; in these circumstances it is obviously impossible to get an end-point.Apparently the lower cyanide of cobalt first formed withdraws cyanide from the soluble silver cyanide complex, itself forming a higher cyanide silver iodide being precipitated. This process in itself slow naturally becomes slower as less and less cobalt remains to be converted. Attempts to speed up the conversion were not encouraging; heating accelerated it considerably but as cyanide is very apt to react with ammonia and the titration has to be conducted in its presence the liquid began to turn brown owing to the decomposition of the cyanide long before the cobalt was all converted.The first step necessary before one could add the excess of cyanide required for the conversion was to find some means of rendering the solution alkaline without using ammonia; sodium hydroxide was not satisfactory as the precipitate formed did not readily dissolve in the cyanide; sodium carbonate was better but by no means perfect; sodium bicarbonate better still. Eventually I found that an excess of bora 366 EVANS THE DETERMINATION OF COBALT gives a pink precipitate which can be approximately titrated away with cyanide without using silver at all. I t is necessary to run the borax soldtion in all at once; if it is added gradually a blue precipitate which is not readily soluble in cyanide, tends to be formed.A number of experiments were next tried to find out the best method of con-verting the cobalt to the higher cyanide; the requirements were these :-(a) com-plete conversion; ( b ) a stable end-product; (c) no loss of cyanide either by destruc-tion or volatilisation. Being at that time under the impression that cobalticyanide was the end-product to be aimed at I attempted the conversion by adding a variety of oxidising agents as well as the excess cyanide; the figures obtained fell roughly into two groups: (i) With mild oxidising agents e.g. air bubbling even at boiling temperature, the consumption of cyanide was considerably less than that required for the Co (CN) ratio. (ii) With more drastic treatment e.g. boiling with hydrogen peroxide the consumption of cyanide was much greater than that calculated and cyanide had obviously been destroyed.On examination I found that the figures obtained with the milder treatments tended to approximate to those required for a Co:(CN) ratio which agrees with Glasstone and Speakman’s finding; but the agreement was by no means perfect; even with the mildest treatment a decided excess of cyanide over that ratio had been used up and a considerable increase in the severity of treatment did not increase this excess by very much. This is illustrated by the following figures: Cobalt taken . . . . 0.0050 g. No. of ml. KCN solution required for Co (CN) . . . . 6.23 ml. ) ) > > J ) > > I 9 , Co:(CN) . . 7.46 ,, 9 , , t , , air-bubbled for 5 minutes cold 6-40 ,, J ) J J I ) > ) 2 ) heated to boiling removed from plate air-bubbled 5 minutes 6.45 ,, A search in the literature of the subject revealed the statement by Manchot and Herzog28 that hydrogen peroxide is set free during the atmospheric oxidation of potassium cobaltocyanide.This gave an explanation of the anomalous results, and I then found that if the potassium iodide which has in any case to be used as an indicator in the final titration is added before the initial titration with cyanide, results agreeing closely with the Co (CN) ratio are obtained. The following process was worked out: PROPOSED VOLUMETRIC METHOD.-SOZ&~O~ZS Required. Potassium cyanide (standard) 16.8 g. of KCN with approx. 10 g. of NaOH Silver nitrate (standard) 5.792 g. of pure AgNO per litre.Borax saturated aqueous solution. Potassium iodide 4 per cent. aqueous solution. Sodium carbonate 10 per cent. aqueous solution. Ammonium chloride 20 per cent. aqueous solution. Dilute ammonia (1 1). made up to 3500ml EVANS THE DETERMINATION OF COBALT 367 The faintly acid solution containing the cobalt to be titrated is placed in a flask and diluted to about 100ml. the flask is swirled round and 20ml. of the borax solution are tipped in from a measuring cylinder; this should result in the formation of a pale pink precipitate in the liquid; 10 ml. of the potassium iodide solution are next added and the mixture is titrated with the standard potassium cyanide solution until the precipitate has dissolved and the liquid is only slightly hazy.The volume of cyanide used V is read and an excess is added bringing the total to (1-3V + 5). The exact volume added does not matter so long as there is an excess which is provided for by the formula. As the titration has to be continued from the same burette after a few minutes and as cyanide burette taps are very apt to leak it is advisable to stand a small beaker under the burette while the process is being continued. To the liquid in the flask now containing an excess of cyanide are next added 10ml. of the sodium carbonate solution and a rapid stream of air is drawn or forced through it for six minutes. This can conveniently be done by means of a wash-bottle attachment with the short arm connected with a filter-pump. It is advisable to adhere fairly closely to the time of aeration stated as with less than five minutes the conversion seems to be incomplete whilst with more than seven or eight, results tend gradually to be slightly high.At the end of the aeration the pump is disconnected the attachment is rinsed in and a mixture of 10 ml. of the dilute ammonia and 25 ml. of the ammonium chloride solution is added. The beaker which has been standing under the cyanide burette is rinsed into the flask and the solution which should be ,quite bright is titrated with standard silver nitrate solution until it is permanently opalescent. Titration is now con-tinued from the cyanide burette drop by drop with vigorous shaking until the liquid again becomes bright. This is the end-point and the burette readings are noted. The burettes having been refilled an amount of the standard cyanide 0.5 ml.less than the previous burette reading is run into the solution. This is again titrated to opalescence with the silver solution and again brought back to brightness with the cyanide; the two solutions are then run in alternately drop by drop until with the original amount of cyanide added the end-point is reached. The new silver reading gives the amount of silver solution required to balance the amount of cyanide used the original one that required for the excess cyanide; the difference between the two is a measure of the amount of cobalt combined with the cyanide; this number multiplied by 0.803 gives the weight of cobalt in milligrams. Experimental results were as follows : TABLE I Cobalt taken g.0*05000 0.02992 0.02500 0*02000 0.01497 0~01000 0-00500 0*00200 Tit ra tion Ell. 80*00- 17*70=62.30 49.80- 12*55=37*25 44.10- 13*00=31* 10 38~00-13~00=25*00 28.45- 9*80=18*65 24055- 12.15=12*40 11-85- 5.75s 6.10 10.70- 8*20= 2.50 Cobalt found g-0.05001 0.02992 0.02498 0.02006 0-01498 0.00995 0*00490 0*0020 368 EVANS THE DETERMINATION OF COBALT Titrations carried out on known mixtures of nickel and cobalt showed that, as might be expected the figures for the two metals are merely additive; these results are given in Table 11. TABLE I1 Cobalt Nickel taken taken Titration g- g. ml. Titration calculated ml. 0.0300 0.0030 50.70 - 10*25=40.45 40.40 0.0240 0*0090 48.90- 10*00=38.90 38.90 0.0180 0.0150 48*90- 11*60=37.30 37.40 0.0120 0.0210 47.80 - 11*90=35.90 35.95 0.0060 0.0270 48.60 - 14-05 = 34.65 34.50 - 0.0330 46.00- 12.95=33*05 33-00 ordinary way 0.0330 36.80- 3*75=33*05 33.00 Direct titration in The last two determinations show that the titration of nickel gives the same result by the process as by the ordinary cyanometric method.The factor for nickel is 0.0010. Originally 10 ml. of 20 per cent. sodium hydroxide solution were used instead of the sodium carbonate added after the initial titration. This gave accurate figures with the higher amounts but the results tended to be low with the lower amounts. On the other hand when no alkali at all was used high results were obtained presumably owing to slight decomposition of the excess cyanide.TABLE I11 Sodium Sodium hydroxide carbonate soh tion solution Cobalt Cobalt taken (20 per cent.) (10 per cent.) recovered g. ml. ml. g-0-0250 10 - 0.0249 0.0250 - 10 0.0250 0~0100 10 - 0*0095 0*0100 10 - 0.0099 0~0100 10 - 0.0094 0.0100 - 10 0*0100 0*0100 no alkali 0-0104 TITRATION OF NICKEL IN PRESENCE OF CoBALT.-In view of the recently assumed industrial importance of electro-deposited nickel-cobalt alloys it seemed worth while to try and find a reliable method for the volumetric determination of one or other constituent independently which combined with the joint titration given above would give a complete analysis of the alloy. Not having had very happy experience of the cobaltic hydroxide titration (vide S Z ~ ~ ~ Y I X ) I turned my attention to the conversion of the cobalt into the very stable cobalticyanide form EVANS THE DETERMINATION OF COBALT 369 followed by recovery of the nickel from its more unstable complex.This principle has been used before by Feigl and Kapulit~as,~~ who however add formaldehyde to remove the excess cyanide thus rendering impossible a subsequent direct cyanide titration. The recovery of nickel from its cyanide complex is not quite so simple as at first sight appears. The complex is readily decomposed by acids, but the immediate result is the formation of an extremely insoluble white pre-cipitate-apparently a nickel nickelocyanide analogous to Prussian blue ; this precipitate resists all further attempts at decomposition. Prolonged boiling of the complex with a large excess of very faintly acid ammonium citrate effects de-composition and the nickel can then be titrated; results however though approximately correct seem to be invariably somewhat low; probably still a trace of the insoluble compound is formed.The following extremely simple procedure was found to give entirely satisfactory results : The solution is made slightly ammoniacal an excess of potassium cyanide (2-3 ml. of the saturated solution should be ample for an ordinary alloy) is added, followed by a few drops of hydrogen peroxide (20 vol.); the solution is then boiled for 5 minutes. The cobalt should now be entirely converted into the inert cobalticyanide while the nickel remains principally as nickelocyanide ; 20 ml. of ammonium chloride solution (20 per cent.) are added and the liquid is boiled for a further 15 minutes.The flask is now removed from the plate and 10ml. of dilute (1 1) ammonia followed by 10 ml. of hydrogen peroxide are added; the flask is replaced on the hot plate and boiled very gently for 10 minutes. This stage of the process is the only one which needs careful attention; it is necessary to dispel all the hydrogen peroxide but if the ammonia is boiled off too far nickel cobalticyanide comes down as a white precipitate. Nickel cobalticyanide is not readily re-dissolved in a small excess of ammonia; if it has been precipitated the titration can still be performed but it takes longer. The flask is now cooled, a few drops of citric acid solution (100 g. to 200 ml. of water) are added followed by 10 ml.of dilute (1 :1) ammonia and 10 ml. of potassium iodide solution (4 per cent.) and the solution is diluted to approximately 250-300 ml. with water. The titration is carried out as follows:-Standard potassium cyanide is run in until the blue colour of the solution is practically dispelled then 1 to 2 ml. of the standard silver solution are added from the burette. This probably produces a turbidity, and in that case more cyanide is run in until the turbidity is dispelled and the solution is bright. If nickel cobalticyanide has been precipitated it will still be visible at this stage and if so a few ml. excess of potassium cyanide solution must be added and the flask allowed to stand with occasional shaking until all solid particles are dissolved. In either case the turbidity is brought back by the addition of a little more silver nitrate and then just dispelled again by the addition of potassium cyanide added drop by drop.The silver nitrate value of the total potassium cyanide added is found in exactly the same way as in the cobalt titration by adding nearly the full amount to the titrated liquid; titrating this away with silver nitrate and then cautiously titrating alternately with the two solutions until the balance occurs when the full amount of potassium cyanide has been added. The difference in ml. between the two volumes of standard silver nitrate solution thus obtained is equal to the weight in milligrams of nickel present 370 EVANS THE DETERMINATION OF COBALT The following experimental results were obtained by this method : TABLE IV Taken 7 Nickel Titration found g .ml . g -P- Cobalt g. 0.0300 0.0030 35.60 -5.60=30.00 0.03000 0.0270 0*0060 36.00 - 9.00=27*00 0*02700 0.0240 0.0090 31.65 - 7.65=24*00 0.02400 0.0210 0.0120 24.40 - 3*50=20.90 0-02090 0.0 180 0.0 150 21.35 - 3.40 = 17.95 0.0 1795 0.0150 0.0180 17.70 -2*80= 14.90 0.01490 0-0120 0.0210 13.70 - 1.70 = 12-00 0~01200 0*0090 0.0240 12*50-3*45= 9.05 0.00905 0*0060 0.0270 8.25-2*10= 6.15 0.0061 5 0.0030 0*0300 4.50 - 1*50= 3.00 0.00300 The complete process therefore for the determination of nickel and cobalt in a mixture of these two metals gives us two titration figures expressed in ml. of the silver nitrate solution (i) combined nickel + cobalt titration = x ml. (ii) nickel titration = y ml.then: The method was tried on five electro-deposited alloys of nickel and cobalt supposedly free from other metals. Cobalt = ( x - y ) 0.803 mg. Nickel = y mg. Results were as follows: TABLE V Weight taken Combined f& each titration g . 0-0320 0.0320 0.0320 0.0320 0.0320 titration (ml. AgNO,) 34-05 34-45 35.45 36.15 38.75 Nickel titration (ml. AgNO,) 22.90 20.75 17.20 14.00 4.80 Cobalt found g. 0.00895 0-01 100 0.0 1466 0.01778 0.02725 Nickel found €5 0.02290 0.02075 0.01720 0.01400 0.00480 Composition per cent. CO 27.98 Ni 71.58 99.56 CO 34-38 Ni 64.85 99.23 CO 45.80 Ni 53-73 99.53 co 55-59 Ni 43.76 99.35 CO 85-12 Ni 14.98 100~11 -= -= EVANS THE DETERMINATION OF COBALT 371 It must be clearly borne in mind that before the cobalt titration can be carried out other metals especially iron copper and zinc must be eliminated.far the most widely used method for the separation of cobalt from most metals appears to be its precipitation by a-nitroso-IS-naphthol. It is especially un-fortunate that iron is also precipitated by this reagent in view of the fact that this metal almost always accompanies cobalt that many industrial alloys of cobalt also contain notable amounts of iron ( e g . cobalt steels) and that really reliable and simple methods of eliminating iron which do not involve filtration of large amounts of ferric hydroxide are conspicuously lacking (vide supra). The most usual way of removing the iron preliminary to determining cobalt in steel analysis is by zinc oxide separation.The objections to this method have been stated earlier (vide supra). Both the 8-hydroxyquinoline and the pyridine thiocyanate methods precipitate iron as well as cobalt. Many methods have been published which claim to effect the ~ e p a r a t i o n ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ but all or almost all of these methods, granted that they are all that is claimed for them suffer from the serious dis-advantage (from the point of view of steel analysis) that it is the iron which is precipitated whilst the cobalt remains in solution. The claim made for the process which follows is that the cobalt is precipitated substantially free from most metals including iron without any preliminary separation what ever.The usual solvent used for a-nitroso-IS-naphthol is acetic acid and this solution as has been said precipitates iron as well as cobalt. The reagent is, however soluble in a number of other liquids and in syrupy phosphoric acid it gives a solution which while still precipitating cobalt copper and ferrous iron,= no longer precipitates ferric iron. The ordinary medium in which precipitation is effected is dilute hydrochloric acid This may however be replaced by dilute nitric acid without ill effect so long as the nitric acid is free from oxides of nitrogen. This fact enables alloys to be readily brought into solution with the iron in the ferric condition without the introduction of possibly harmful reagents. One difficulty in the a-nitroso-p-naphthol precipitation has always been that the reagent itself is precipitated as well as the cobalt complex.This makes it difficult to know when an adequate excess of the precipitant has been added; the technique adopted in the process to be described overcomes to some extent this difficulty. A double precipitation has been adopted; this is principally due to the fact that the nitroso-naphthol precipitate is very bulky and difficult to wash and as a matter of fact has always been the correct procedure where large amounts of metal ( e g . nickel) in solution have to be eliminated. Copper still precipitates with the cobalt and as it reacts with cyanide must be removed; this is achieved by a hydrogen sulphide precipitation before the second nitroso-naphthol precipitation ; as pointed out earlier there is the possibility that cobalt may tend to co-precipitate with copper sulphide.The amount of copper in steels however is generally so small that the co-precipitation may be ignored; with an alloy containing copper it will probably be desirable to re-precipitate the sulphide. Attempts were originally made to determine the cobalt gravimetrically. These were unsuccessful as the results were invariably somewhat high; apparently it is difficult to remove all SEPARATION OF COBALT FROM IRON ETC. ; DETERMINATION I N STEEL.-B 372 EVANS THE DETERMINATION OF COBALT phosphoric acid from the precipitate but in view of the success of the volumetric process this question has lost its significance. The process finally worked out for the determination of cobalt in steel is as follows: DETERMINATION OF COBALT IN STEEL.-The sample (5g.for low amounts, less for higher) is dissolved in dilute nitric acid (spgr. 1.2) in the proportion of 75 ml. per 5 g.; the solution is boiled for 10 minutes and cooled and about 0.5 g. of urea is added. The reagent solution which should not be more than 2 or 3 days old is prepared by placing the wnitroso-j3-naphthol in a small beaker adding syrupy phosphoric acid in the proportion of 50 ml. to 3.5 g. of nitroso-naphthol, placing the mixture on the hot plate and stirring it with a thermometer until the temperature reaches 60" C. and finally cooling. The solution of the sample is diluted to about 100 ml. and the reagent solution is added 1 or 2 ml. at a time, with subsequent stirring a few seconds being allowed to elapse between additions during which stirring is continued.The following phenomena are observed The liquid darkens then a red turbidity which increases as the additions continue is produced and after two or three additions a stable froth is formed and carries up the red precipitate. At a certain point with continued addition and stirring the precipitate suddenly begins to darken the froth breaks or shows a tendency to break and a black resin-like scum floats to the surface. This marks the presence of an excess of reagent but it is advisable to add a final quantity of 3 to 4 ml. The liquid is next diluted with about its own volume of 20 per cent. acetic acid and stirred and about 3 ml. more of the reagent are run in (omission of this last precaution seems for some reason to lead to a tendency to slightly low results).The liquid is allowed to stand (cold) for about 25 minutes (not less than 20) with occasional stirring and the precipitate is filtered off on paper and well washed with cold dilute nitric acid (5 per cent. prepared by diluting 20 per cent. nitric acid which has been boiled for 15 minutes). . The precipitate so obtained is dark coloured probably nearly black and very bulky and may contain a good deal of water. The filter and its contents are transferred to a roomy platinum dish and placed on the plate to dry. (This is usually advisable as quite a considerable amount of liquid may separate on heating.) When sufficiently dry the dish is placed in the muffle and the filter is burnt off at not too high a temperature.If vanadium molybdenum or tungsten (vide inf~a) is present the residue is fused with fusion mixture taken up with hot water and filtered the precipitate is washed with hot water and the filter is replaced in the platinum dish and again burnt off; in the absence of these elements this step may be omitted. When completely burnt off the residue is fused with potassium bisulphate heating being continued until potassium sulphate begins to crystallise as a scum on top of the melt. If an appreciable amount of cobalt is present the melt will be dark blue when hot and pink when cold; in absence of copper it may be dissolved in about 100 ml. of 5 per cent. nitric acid the solution boiled and cooled and the second precipitation carried out forthwith but with a steel it is safer to proceed as follows: The melt is dissolved in water and the solution is transferred to a beaker, hydrogen sulphide is passed the liquid is heated to boiling and the precipitate is allowed to settle then filtered off and washed with hot water.The filtrate i EVANS THE DETERMINATION OF COBALT 373 boiled until hydrogen sulphide has been dispelled 25 ml. of the boiled-out 20 per cent. nitric acid are added and the liquid is boiled down to a volume of about 100 ml. After cooling the second precipitation is effected in exactly the same way as the first but no urea is added. The resulting precipitate should be crimson lake in colour. It is desirable to test the filtrates of both this and the first precipitation for excess of reagent; this may be done by removing about 50ml.of the filtrate (before washing) dividing it into equal parts and adding about 0.2 mg. of cobalt to one of them. After standing for some time the formation of the cobalt precipitate in one of them generally causes a distinct difference in appearance between the two parts. Too much stress must not be laid on this test; with the filtrate from the second pre-cipitation it is almost invariably and immediately successful but in the first precipitation the excess of the reagent seems itself to be mainly precipitated or destroyed. After washing the filter and precipitate are transferred this time to a porcelain crucible. Platinum must not be used as the subsequent solution of the cobaltic oxide in hydrochloric acid liberates chlorine; silica does not seem so satisfactory as porcelain.The filter will probably need drying again before ignition; it is then burnt off as before care being taken to see that the burning off is complete and that no carbon particles remain. After ignition the residue should be black; it may have white or blue patches presumably due to phosphoric acid which has escaped removal. When the residue is cool about 10 ml. of strong hydrochloric acid are poured on part of it being used to rinse down the sides. The crucible is then covered with a watch-glass and stood on a steam-bath. When the acid is hot a fairly vigorous reaction ensues which should result in the residue dissolving completely to a blue solution. When this is achieved the cover-glass is removed and rinsed in with hot water and the acid liquid is evaporated to dryness over steam.The sides of the crucible are rinsed down with about 15 drops of strong nitric acid, the solution is again evaporated to dryness and the residue of cobalt nitrate (which should be dark red) is taken up with 2 drops of nitric acid and about 10 ml. of hot water; this should give a pink solution which is probably slightly hazy. The solution is transferred to a beaker and cooled; it may contain a very small trace of iron which must be removed not because it is sufficient to influence the titration, but because it may cause a haze in the liquid and make the end-point difficult to observe. The removal is effected by the addition to ihe cold liquid of a suspension of barium carbonate in water added a few drops at a time until a trace of undissolved barium carbonate is floating in the solution as a semi-flocculent precipitate.* The precipitate along with any trace of iron that was present is filtered off and the precipitate is washed with cold water.The filtrate is now ready for the addition of borax and the remainder of the volumetric procedure is carried out as already described. Only one slight modification has to be introduced in this It is allowed to settle filtered off and washed as before. * Great care must be taken to avoid any excess of barium carbonate otherwise a little cobalt may be lost by adsorption (vide infra) 374 EVANS THE DETERMINATION OF COBALT procedure because of the soluble barium salts produced by the barium carbonate treatment.When sodium carbonate is added after the initial titration this of course precipitates the barium; therefore after aeration and the subsequent addition of ammonia and ammonium chloride the liquid is filtered and the pre-cipitate is washed with cold water. The final titration is then proceeded with as usual. The trace of iron requiring removal by barium carbonate is so small that no appreciable amount of cobalt is carried down with it. TABLE VI Trials of this method were carried through with the following results: Per cent. of cobalt Iron taken g. 5.00 5.00 5.00 5.00 5.00 5.00 5.00 Nickel Copper taken taken €!* g-0.050 0.050 0.050 0.050 0.050 0.050 0.050 0.010 Cobalt taken g. Blank 0*0020 0.0050 0~0100 0.0250 0.0500 0~0100 Titration ml.of AgNO soh. 3.40- 3.25= 0.15 7.65- 5.10= 2.55 16.35- 10*00= 6.35 24.90- 12.70=12*20 38.85- 7*55=31.30 77.80 - 15.90 = 61 -90 31-20 - 19.10 = 12.10 Cobalt found g. 0~00012" 0.00205 0.00510 0*00980 0.02513 0-04972 0.00972 f 1 found (corr. for added blank) - -0-0400 0-0386 0.1000 0.0996 0.2000 0.1960 0*5000 0.5002 1*0000 0.9920 0.2000 0.1920 * That this blank was really due to cobalt was shown by the colour of the hydrochloric acid extract of the residue. The method was tried without modification on steel "W" of Messrs. Ridsdale's The percentage composition of this steel as given in C 0.695; Si 0.187; S 0.075; P 0.028; Mn 0.101; Cr 3-01; V 0.791; As I was then unaware of the precipitation of vanadium etc.by the reagent, the first precipitate was fused direct with bisulphate the fusion with fusion mixture being omitted. The very considerable amount of tungsten (16.21 per cent.) was ignored throughout and was finally filtered off with the barium carbonate precipitate just before the titration was begun. British Chemical Standards. the certificate is: W 16.21; Co,'4.76; Ni 0.44; As 0.01; Cu 0.058; Mo 0.048. The results obtained were as follows: TABLE VII Weight taken Titration €5 ml. Cobalt Per Cent. 0.5000 47.20-18*70=28*50 4.58 0*5000 47.30- 19*00=28*30 4.55 0.5000 43.75- 15*30=28.45 4.57 0*5000 43.45 - 15*30=28*15 4.52 The cobalt figure 4.76 given in the certificate is the mean of a large number of determinations by 19 independent referees using a great variety of methods; the individual figures given vary from 4.52 to 4.99.As I made no attempt to work u EVANS THE DETERMINATION OF COBALT 375 the barium carbonate precipitate (vide infra) it is very probable that my figures are slightly low. Interference by Chromium Vanadium etc.-After completing the work so far described I came across the statement in Hillebrand and Lundell’s text-book:’ that bismuth silver chromium zirconium titanium vanadium tin and nitric acid* also “interfere” (with the gravimetric determination of cobalt by precipitation with a-nitroso-/3-naphthol). Time did not permit of an exhaustive examination of this question but in view of the use of chromium vanadium and titanium in d o y steels it was necessary to ascertain the effect of these metals at least on the process.As is proved by the figures given in this paper nitric acid is entirely harmless if certain simple precautions are taken; further experiments showed that : ( a ) Chromic salts do not seem to be precipitated by the reagent to any extent; a slight precipitation takes place but this is presumably due to adsorption, and possibly also to the fact that ( b ) Chromic acid appears to be precipitated. (c) Vanadic acid is precipitated (vanadyl salts were not tried). ( d ) Vanadic acid does not interfere with the cyanide titration. In the process as applied to steel the chromium is left as chromic salt; the vanadium on the other hand is largely converted into vanadic acid; consequently, in the first precipitate one has ( a ) All the cobalt; ( b ) all the copper; ( c ) at least a great part probably the whole of the vanadium molybdenum tungsten etc.; ( d ) a little chromium and iron. Of these the cobalt and copper have already been dealt with; the trace of chromium behaves in much the same way as a trace of iron and is eliminated therewith; such of the vanadium etc. as gets through into the titration liquid exerts no influence and in any case vanadium tungsten and molybdenum are removed by the fusion with fusion mixture. An initial experiment (made before I realised the extent to which vanadium is precipitated) was carried out with a solution containing 1.452 g. of chromium and 0.700 g. of vanadium. This was divided into two equal portions and 0.010 g.of cobalt was added to one of them. After adjustment of the nitric acid strength the two solutions were precipitated with the phosphoric acid-nitroso-naphthol reagent in the usual way; the precipitates obtained were burnt off in porcelain, and the residues were taken up with hydrochloric acid thus proceeding direct to the latter part of the process without a second precipitation. A great deal of vanadic acid did not go into solution when the crucible was rinsed out and was consequently filtered off with the barium carbonate precipitate. The titration behaved normally but the result was somewhat low-0.0090 g. of cobalt in place of the 0.01OOg. taken. The barium carbonate precipitate on being worked up (vide i n f y a ) furnished a further 0.0005 g.which had been entangled in the un-dissolved vanadic acid giving a total recovery of 0.0095 g.; the blank required 0-30 ml. which quantity was deducted from the titration figure. The lost 0.0005 g. * Lundell Hoffman and BrightS8 say “Iron copper molybdenum chromium zirconium, titanium vanadium tin and tungsten are also precipitated. Nitric acid interferes. 376 EVANS THE DETERMINATION OF COBALT was probably not precipitated owing to an insufficient quantity of reagent having been used for so large an amount of vanadium. The second experiment was carried out with smaller quantities of vanadium and chromium but with addition of titanium; each solution contained 0.025 g. V, 0-037 g. Ti and 0-145 g. Cr and 0.010 g. of cobalt was added to one of them. The process used was as for steel but the first precipitate after being burnt off was fused with fusion mixture.The melt was taken up with water and the precipitate was filtered off and washed with hot water. The precipitate was again burnt off and then fused with bisulphate the resulting melt being treated like the ordinary bisulphate melt of the first precipitate (reprecipitation etc.). Copper being absent hydrogen sulphide was not used. The titration of the cobalt sample required 11-65 ml. and that of the blank 0.15 ml.; the cobalt corresponding to the difference of these figures is 0.00924 g. There had therefore been a loss of 0-00076 g. , or 0.95 ml. The barium carbonate precipitate was then worked up (vide ircfra), and the titration resulting was 0.70 ml. bringing the total cobalt recovered up to 0.00980 g.It would seem therefore that the trace of chromium (or vanadium?) left after the second precipitation tends to drag down a fraction of a milligram of cobalt in the barium carbonate precipitate. It is clear that none of the three metals tested in any way prevents the pre-cipitation of the cobalt and the only form of interference shown was the removal of a trace in the barium carbonate precipitate. For ordinary purposes this trace can probably be neglected; for accurate work it can be recovered as follows:-The barium carbonate precipitate is burnt off in a platinum dish and fused with potassium bisulphate the melt is taken up with 5 per cent. nitric acid boiled and filtered the filtrate is then treated with the nitroso-naphthol reagent in the same way as the solution of the bisulphate melt of the first precipitate in the ordinary process and the determination is finished as usual.In the above trial about a milligram of blackish residue was obtained on burning off this precipitate and this dissolved in hydrochloric acid to give a blue solution; the hydrochloric acid solution of the “blank” residue was not blue and the 0.15 ml. titration figure was probably an experimental error. The vanadium extracted in the alkali fusion of the first precipitates was titrated; that from the cobalt sample contained 0.0135 g. and that from the blank 0-0158 g. out of a total in each case of 0.0250 g. The loss seems to have occurred during burning off. I hope to investigate the vanadium question later with a view to a possible method for determining vanadium in steel.Lack of time prevented further work being done on the separation of cobalt from chromium vanadium titanium, molybdenum etc. but as indicated above there seems no reason to fear that precipitation will be incomplete or that alkali fusion will fail to remove the vanadium molybdenum etc. It might be mentioned in conclusion that precipitation of cobalt in the usual way with an acetic solution of the nitroso-naphthol and with phosphoric acid added to the cobalt solution does not have the desired effect iron being precipitated to a considerable extent ; increase in the amount of phosphoric acid with the object of holding the iron up completely seems to prevent the complete precipitation of the cobalt.Several experiments carried out with the aim of preventing th EVANS THE DETERMINATION OF COBALT 377 reagent precipitating while allowing of the complete precipitation of the cobalt complex by means of the addition of excess of acetic acid acetone etc. appeared to indicate that under these conditions the solubility of the cobalt complex is by no means negligible. The cobalt solution used for these experiments was made up from a specially prepared sample of very pure electrolytic cobalt; the iron used was electrolytic iron of a high degree of purity; the nickel was a very pure sample of Mond nickel, which previous analysis had shown to contain 0.02 per cent. of cobalt. All reagents used were of analytical reagent quality except the a-nitroso-g-naphthol, which was old stock.SUMMARY.-(&) A new volumetric method of determining cobalt has been described. This method closely resembles the ordinary nickel titration. ( b ) A method has been described for titrating nickel in the presence of cobalt. (c) A new method of separation of cobalt from all metals interfering with (a) A preliminary survey of other metals precipitated by the reagent has the volumetric procedure has been detailed. been made and their non-interference with the method has been demonstrated. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. REFERENCES D. H. Brophy Ind. Eng. Chem. L4nal.Ed. 1931 3 363. W. R. Schoeller and A. R. Powell J . Iron and Steel Inst. 1918,97 444; ANALYST 1919,44, Recent Advances in Analytical Chemistry Vol. 11 pp. 266-7. H. H. Willard and D. Hall J . Amer. Chem. SOC. 1922 44 2219 2226 2237. A. A. Guntz and J. Barbier Chem. et Ind. 1929 21 711. R. Berg Z. anal. Chem. 1929 76 191. H. R. Fleck and A. M. Ward ANALYST 1933 58 388. M. Auger and L. Odinot Compt. rend. 1924 178 710. W. Funk 2. anal. Chem. 1907 46 93. W. Bottger and K. Druschke Annalen 1927 453 315. E. Swift and R. Barton J . Amer. Chem. Soc. 1932 54 2220. M. Ilinski and G. von Knorre Z. angew. Chem. 1893 264. J. I. Hoffman Bureau of Standards J . of Research 1931 7 883. W. R. Orndorff and M. L. Nichols J . Amer. Chewz. SOC. 1923 45 1439. E. Muller and H. Lauterbach Z. anal. Chem. 1923 62 23. L. A. Sarver Ind. Eng. Chem. Anal. Ed. 1933 5 275. G. Barbieri Atti. R. Accad. Lincei 1928 [v] 8 405. A. A. Wassilief Z. anal. Chem. 1929 78 439. H. Yagoda and H. M. Portridge J . Awzer. Chem. SOC. 1930 52 4887. G. Spacu and M. Kuras Bul. SOC. Stunte Cluj 1934,7 377; BY. Chem Abst. 1934 1323. P. Spacu Compt. rend. 1935 200 1595. J. T. Dobbins and J. P. Saunders Ind. Eng. Chem. Anal. Ed. 1934 6 459. J. Ledru and L. Hauss Bull. SOC. Chim. Belg. 1932 41 104. R. Uzel*and B. Jezek Coll. Czech. Comm. 1935 7 497; Br. Chew Abst. 1936 444. S. Glasstone and J. C. Speakman ANALYST 1930 55 93. E. Rupp and F. Pfenning Chew-Ztg. 1910 321, W. Manchot and J. Herzog Ber. 1900 33 1742. F. Feigl and H. J. Kapulitzas,,Z. anal Chem. 1930 82 417. F. Zeitsche and M. Nachmann Helv. Chim. Acta 1926 9 420 705 979. S. Krishna and H. Singh J . Amer. Chem. SOC. 1928,50 792. P. Ray and A. Chattopadhya 2. anorg. Chem. 1928 169 99. G. Font& and L. Thivolle Bull. SOC. Chim. 1924 35 641. C. Mayr and F. Feigl 2. anal. Ckem. 1932 90 15. M. Ilinski and G. v. Knorre Ber. 1885 18 699. L. Dufty J . Iron and Steel Inst. 1914 11 p. 52. Hillebrand and Lundell Applied Inorganic Analysis (1929) p. 326. Lundell Hoffman and Bright Chemical Analysis of Iron and Steel (1931) p. 337. 279. - Id. 1932 8 659. - Id. pp. 337-338

ISSN:0003-2654    

DOI:10.1039/AN9376200363

出版商:RSC    

年代:1937

数据来源: RSC

摘要:

OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc.it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned.In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix.about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary.Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary.It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order.It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained.In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents. To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'.It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air.There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary.It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined.It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'.It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place.The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C.It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined.It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'.It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place.The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix.about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary.Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,

ISSN:0003-2654    

DOI:10.1039/AN9376200378

出版商:RSC    

年代:1937

数据来源: RSC

摘要:

OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc.it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned.In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix.about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary.Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary.It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order.It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained.In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents. To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'.It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air.There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary.It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined.It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'.It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place.The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C.It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined.It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,

ISSN:0003-2654    

DOI:10.1039/AN9376200383

出版商:RSC    

年代:1937

数据来源: RSC

摘要:

OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc.it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned.In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix.about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary.Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary.It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order.It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained.In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents. To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'.It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,

ISSN:0003-2654    

DOI:10.1039/AN9376200387

出版商:RSC    

年代:1937

数据来源: RSC

摘要:

OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc.it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned.In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix.about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary.Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary.It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order.It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained.In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents. To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'.It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,

ISSN:0003-2654    

DOI:10.1039/AN9376200389

出版商:RSC    

年代:1937

数据来源: RSC

摘要:

OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc.it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned.In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix.about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary.Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary.It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order.It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc. it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained.In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents. To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'.It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions. The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,

ISSN:0003-2654    

DOI:10.1039/AN9376200391

出版商:RSC    

年代:1937

数据来源: RSC

摘要:

OF MILK: CORRECTION FACTORS AND THE INFLUENCE OF STIRRING: I1 h I11 225 It was originally intended to employ four different amounts of supercooling for each sample of milk, vix. about 1.5, 1.0,0.8, and 0.5" C. It was found, however, that the differences in the readings of the thermometer corresponding to these different amounts of supercooling were in the neighbourhood of 0.002"-an amount so small that the unavoidable errors of observation might possibly approach the same order. It was, therefore, decided to employ only two different amounts of supercooling, as widely apart as practicable, say about 1.5" and 04", and to increase the number of samples examined. It was stated by Hortvet that, unless a much greater amount of supercooling than 0.5" is employed in the use of his cryoscope and technique, the rise of the mercury column is not sufficiently pronounced, and that there is more or less wavering, so that difficulty arises in deciding on the exact point at which the top of the column becomes stationary. Elsdon and Stubbs (Eoc.it.)^ found the same result when using a supercooling of less than about 0.8"; the mercury rose very slowly, and did not maintain a steady position for any appreciable time, and the proper freezing-point might not, under these circumstances, be attained. In the first experiments, made with the object of ascertaining the super- cooling correction, it was observed that where the amount of supercooling was small, that is, less than say 0-75", even in the absence of alcohol in the jacket surrounding the freezing-tube, the rise of temperature, when freezing occurred, was very slow-so slow, indeed, as to suggest doubts whether the thermometer would indicate the freezing-point of the milk, influenced only by supercooling, owing to the reading being affected by the length of time which elapses and the possibility of imperfect thermal insulation, causing a nett loss of heat from the freezing tube and contents.To give an instance; in an experiment when the supercooling was 0-71" the time required for the mercury column to rise until, on observation through the telescope with the aid of the horizontal cross-wire, the ascent became imperceptible, was 92 minutes, as compared with about 3& minutes when the same milk was super- cooled 1-49'. It will be seen later that, in the absence of alcohol in the space surrounding the freezing-tube, the heat insulation of the milk is not perfect; a nett loss of heat occurs, for it is possible t o carry through a freezing-point determination under such conditions.The difference between the two methods of working- with and without alcohol-results in a longer time being required for the cooling of the milk when the space around the freezing-tube is occupied by air. There were also the difficulties of judging when the rising column of mercury had reached the highest point, due to the very slow movement which occurred in that region, and of getting satisfactory readings, owing to the short time that elapsed before a fall took place. The idea of following strictly the Hortvet technique, as regards stirring, for these particular experiments was therefore abandoned. In the first series of experiments, the results of which are recorded in Table 111, four or five stirrings of three strokes each were employed, and in the second series the mechanical stirrer, working at the rate of 40 strokes per minute, was in operation all the time, the thermometer rising until tapping became necessary. It does not appear that these variations in stirring caused any significant differences in the results obtained, but the times of rising of the mercury, after two different extents, large and small,

ISSN:0003-2654    

DOI:10.1039/AN9376200393

出版商:RSC    

年代:1937

数据来源: RSC