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/AN9366100287

出版商:RSC    

年代:1936

数据来源: RSC

摘要:

288 TEA AND COFFEE THEIR ALKALOIDS AND TANNINS Tea and Coffee with Special Reference to their Alkaloids and Tannins THE following papers were read at the Joint Meeting of the Society with the Food Group of the Society of Chemical Industry on February Gth 1936: THE CONSTITUTION OF TANNINS INCLUDING THOSE OF TEA AND COFFEE BY PETER MAITLAND B.Sc. PH.D. The tannins are a class of amorphous rarely crystalline substances which occur widely in nature and possess the property of changing hide into leather. They are remarkable for their astringent taste and for their many precipita-tion reactions with lime lead acetate alkaloids gelatin albumin and other proteins and also for their colour reactions with iron salts. There have been many attempts to classify tannins and the best and simplest of these is that of Freudenberg who divided them into (i) hydrolysable tannins; (ii) condensed tannins; (iii) unclassified tannins.In spite of the enormous amount of work done upon tannins most of them unfortunately belong to class (iii) and this is due to their amorphous and colloidal nature which makes exact investigation difficult. The tannins which have been selected for this brief survey are shown in the following scheme: (1) Hydrolysable { d { :; (2) Condensed Turkish gallotannin Chinese gallotannin Coffee tannin Ellagic acid tannins Cat echin tannin Quebracho tannin (g) Tea tannin GROUP I. HYDROLYSABLE TANNINS (a) Tzcrkish Gallotannid E d Fischer suggested that this tannin was probably a pentagalloyl-glucose as in formula I (R = galloyl), Turkish gallotannin is obtained from Aleppo galls.I CHOR I I I I I HCOR ROCH HCOR HCO-H,COR R=- CO <-2$ OH Galloyl OH MAITLAND THE CONSTITUTION OF TAXSINS 289 but recognised that this was an ideal formula. As a result of recent work Karrer has put forward the view that this tannin is a mixture of several glucose derivatives, of which the three formulae (11) below are types. H 1- CO . Ellag 1 HyO.Gall I I H?oH :-OCH I H,CO . Gall. Gall H 1- 7OsGal1 HCO .Gall HCO .Gall. Gall HCO . Gall. Gall I I t I '- OCH H,CO Gall (Ellag = ellagic acid Gall = gallic acid) I1 H I- CO . Gall HCO . Gall. Gall HCOH HCOH I I I I - OCH H,CO .Gall. Gall (b) Chinese Gallotannin2 Fischer suggested that it was a penta-digalloyl glucose but realised that this also was an ideal formula.Freudenberg has suggested that the experimental evidence shows the tannin to be a complex mixture of a great variety of (probably) nona-galloylated a- and /3-glucoses the limits of the possibilities of the arrangement of the gallic acid residues lying between formulae I11 and IV. This tannin is obtained from Chinese galls. H -CO-Gall HCO-Gall . Gall HCO-Gall . Gall HCO-Gall . Gall 1 I I I -0CH I H,CO-Gall . Gall I11 H -CO .Gall HCO . Gall HCO . Gall HCO . Gall 1 I I I I -0CH H,CO . Gall. Gall. Gall. Gall. Gall IV -4 penta-digalloyl glucose (I11 + one more Gall.) was actually synthesised by Fischer and Bergmann and this synthetic product and its methyl and acetyl derivatives were shown to be very similar to the natural product and its derivatives.Quite a different type of formula has been brought forward by Nierenstein3 for Chinese gallotannin (V). (OH) . C& . co . 0 . (OH) . C,H,. co I [CO .C,H,. (OH) .O .CO .C,H,. (OH) .O]x I 0. (OH) .C,H,.CHOH.0.C6HB. (OH), 0 I u LO 290 TEA ASD COFFEE THEIR ALKALOIDS .4XD TANNINS He suggests that the galloyl residues are all linked and that the above substance sometimes occurs as a glucoside the glucose being attached to the position marked a. In support of this Nierenstein has shown that the methylated tannin on hydrolysis gives a methylated glucose whereas a compound of Fischer and Freudenberg's formula if methylated and hydrolysed should give free glucose.(c) Cofee Tannin4 The tannin in coffee yields on hydrolysis caffeic acid (VI) and quinic acid (VII) and a residue the constitution of which is not known. The two acids from the hydrolysis have been shown to arise from the presence of chlorogenic acid (VIII) in the coffee. HO COOH OH 0"" LH=CH COOH VI VII The actual tannin may be a derivative of chlorogenic acid or it may be a mixture of several substances amongst them chlorogenic acid. VIII (d) Ellagic Acid Tannins Many tannins on hydrolysis yield ellagic acid (IX) which is derived from The ellagic acid two molecules of gallic acid by oxidation and condensation. co 0 H<-Cz$-YOH IX HO 0 oc tannins are therefore probably variously substituted ellagic acids the substituent groups being linked to the ring by means of the phenolic oxygen atoms NAITLAND THE CONSTITUTION OF TANNINS 291 GROUP 11.CONDENSED TANNINS The condensed tannins are either unaffected by acids or they are merely polymerised further. According to Freudenberg many of them are derived from simple crystalline substances called catechins some of which have been found in nature. These catechins are related to the yellow flavanols and the highly coloured anthocyanidin salts which also occur in nature either free or in the form OH HO c1 OH 6H H X XI of derivatives. cyanidin chloride (XI). Thus ordinary catechin ( M I ) is related to quercetin (X) and OH W / C H OH OH H, XI1 ( e ) Catechin Tannin Catechin itself a crystalline compound is not a tannin but by heating it with dilute acid or alkali in the presence or absence of air or by the action of enzymes it is easily polymerised to a substance which will tan hides.In many of the trees in which catechin occurs there is found also a tanning material. This has probably resulted from the polymerisation of catechin but the exact con-nection between this naturally occurring tannin and the synthetic polymerised catechin has not yet been established. (f) Quebracho Tannin Quebracho tannin is obtained from the wood of Quebracho Colorado trees, which are found in the northern part of the Argentine Republic and also in the Gran Chaco. It is one of the most abundant and widely used tanning materials known. In 1925 Freudenberg surveyed all the evidence as to its constitution then available and suggested that it might be derived from a catechin-like substance.Nierenstein had shown in 1906 that the products of potash fusion were mainly resorcinol (XIII) and protocatechuic acid (XIV) and therefore Freudenberg put forward the formula (XV) for the hypothetical quebracho catechin the OH HOOC< >OH <Po> H I - H O p p i I I - v CHOH CH,’ XI11 XIV S 292 TEA AKD COFFEE THEIR ALKALOIDS AND TAKNIXS supposed stem-substance of the quebracho tannin. Freudenberg and Maitland,” in 1934 therefore synthesised this catechin and polymerised it with dilute acid. They were able to establish a connection between the synthetic and natural quebracho tannin on three grounds (1) Analytical results of the same order; (2) the same degradation products; (3) the same type of condensation t o ‘ I phlobaphenes ” without elimination of the elements of water.A tentative suggestion was put forward for the mode of linking of the catechin units in the molecule of the synthetic tannin. This was based on analyses and a degradation reaction. A simple two-unit molecule of the type (XVI) would have the same C and H values as the crystalline quebracho catechin from which it was derived but a higher acetyl value and this was actually observed. The second oxygen ring can open and the condensation can then proceed further and it is very probable that the synthetic tannin is a high molecular compound with the units joined as in formula (XVI). The proof that the link between the molecules is not in the catechol ring A, was found in the observation that oxidation of all both synthetic and natural gave only veratric acid OCH -H O O C C - >CH, XVII The discovery of this completely unexpected type the methylated compounds, (XVII).of condensation without the (XIX) elimination of the elements of water led to a further study of the necessary conditions for a polymerisation of this type. Bergmann and Pojarlieff6 had already made some investigations on this “ phlobaphene reaction.” They first showed that tetramethyl catechin (XVIII) could also undergo the reaction, which proved that the phenolic hydroxyl groups were not necessary for the polymerisation. Examination of the reaction of acids with hydroglucal OCH, H,C CH,O I I I I 0 CH H2cvHoH CHOH XVIII XIS and glucal (XX) showed that the first was stable and the second easily polymerised.They therefore concluded that three essentials were necessary for polymerisation :-(1) Pyran ring; (2) double bond in the pyran ring; (3) hydrosyl in the pyran ring MAITLAKD THE COSSTITUTION OF TANNINS 293 HC CHOH x s Freudenberg and Maitland however prepared (XXI) synthetically and The presence of a hydroxyl in showed that it also was very sensitive to acids. the pyran ring therefore does not seem to be necessary. XXI XXII ( g ) Tea Tannin The evidence for the constitution of this tannin is conflicting some investi-gators having reported it as a " hydrolysable " tannin and others as a " condensed " tannin. An interesting explanation of these divergent views has been brought forward by Tsujimura,' who has suggested that the tannin is a galloyl catechin, in which the galloyl residue replaces a hydrogen atom of the alcoholic hydroxyl I t will be seen from the foregoing brief survey of a few of the well-known tannins that the chemistry of these complex substances is still in its infancy, in spite of the many advances made within the last fifteen years.(XXII). REFEREKES 1. E. Fischer and K. Freudenberg Rer. 1914,47 2485; P. Karrer R. Widmer and M. Staub 2. K. Freudenberg Tannin Cellulose Lagnin p. 38-44. 3. M. Nierenstein The Natural Organic Tannins p. 163; criticised by 0. Schmidt Annalen, 4. K. Freudenberg Bey. 1920 53 232; H. 0. L. Fischer and G. Dangschat Ber 5. K. Freudenberg and P. Maitland Annalen 1934 510 193.6. 7. Annalen 1923 433 288. 1930 479 1. 1932 65 1037. M. Bergmann and G. Pojarlieff CoZZegium 1931 p. 244. M. Tsujimura Sc. Papers I n s f . Phys. and Chem. Research (Tokyo) 1931 15 155. THE UNIVERSITY CAMBRIDG 294 TE.4 AND COFFEE THEIR -4LK.4LOIDS AND TANNINS EXPERIMENTAL WORK ON TEA TANNIN BY M. NIERENSTEIN D.Sc. PH.D. Green Assam tea contains a well-crystallising tannin which is best prepared by the caseinogen method,l in which the amorphous tannin is removed by fractional adsorption. The following formula is provisionally assigned to tea tannin : OR OR A OR lc R = H in the free tannin R = CH in the methyl-derivative OR (i) On methylation with diazomethane a well-crystallising methyl-derivative is obtained which yields on hydrolysis 1 molecule of 6 8 3’ 4’-tetramethyl-l-acacatechin (A) 2 molecules of 3 4-dimethyl-gallic acid (B C) and 1 molecule of trimethyl-gallic acid (D) when hydrolysis takes place at a b and c.(ii) Tea tan.nase obtained by growing Aspergillus niger in a medium containing tea tannin hydrolyses tea tannin at a b and c and yields 1 molecule of Z-acacatechin (A) and 3 molecules of gallic acid (B C D). (iii) Gallotannase obtained from the same mould in a medium of gallotannin,2 hydrolyses tea tannin however only at b and c and yields 3-galloyl-Z-acacatechin (A + B) and gallic acid (C D). (iv) Tannase obtained by growing Aspergillus niger in a medium containing 3-galloyl-Z-acacatechin (A + B) added to gallotannase hydrolyses tea tannin at a b and c ; it thus behaves like tea tannase and produces Z-acacatechin (A) and gallic acid (B C and D).Tea tannase thus consists of two tannases namely 3-galloyl-Z-acacatechin-tannase and gallotannase. REFERENCES 1. 2. - Biochetn. J. 1922 16 614. M. Xierenstein J. Chew. Soc. 1922 121 23. BIOCHEMICAL DEPARTMENT THE UNIVERSITY BRISTO MITCHELL -4 SURVEY OF THE METHODS OF ANALYSIXG TANNINS 295 A SURVEY OF THE METHODS OF ANALYSING TAKNINS BY C. AINSWORTH MITCHELL D.Sc. F.I.C. From its earliest days our Society has interested itself in the methods of determining tannin especially in tea and various gravimetric volumetric and colorimetric methods have first been published in THE ANALYST. PRECIPITATION WITH hkTALLIC COMPOUNDS.-Lead A cetate.-During the first session of the Society Wignerl gave an account of his application of the lead acetate method to tea.I t is not surprising that his results were high ranging from 27 to 45.5 per cent. of tannin. Although it was known that gallic acid and other non-tannins are adsorbed by lead tannate the method found its later supporters. Thus Trillich and Gocke12 used it to determine tannin in coffee a sample of which they found to contain 11.37 per cent. A modification of the method was suggested by Manes the reagent consisting of a mixture of acetic acid and lead acetate solution. Precipitation with Iron Salts.-Handtke4 was the first to suggest the use of ferric acetate as a precipitant for tannin after conversion into sodium tannate. The reagent was also used by Beckmann5 as a substitute for gelatin in the separation of tannin in Lowenthal's permanganate method.This method has the same drawback as the lead acetate method-that of co-precipitating a large proportion of any gallic acid present. To prevent this Ruoss6 precipitated the tannin as a basic tannate in the presence of sodium tartrate and dried the precipitate (to which he attributed the composition C1,H,O,(FeO)) at 120" C. In my experience slight variations in the conditions of precipitation (which with unknown substances cannot be controlled) results in the formation of basic ferric tannate of varying composition. Even complete oxidation with hydrogen peroxide does not always yield concordant results. Other Metallic PreciPitants.-Among the other metallic compounds recom-mended for the precipitation of tannins are potassium antimonate (Gerland'), copper oxide (Flicks) mercuric oxide ( Krugg) and aluminium hydroxide.10 PRECIPITATION WITH GELATIN.-The use of gelatin as a reagent dates back to 1797 when it was shown by Seguinll that tannins form a precipitate with gelatin.Whilst gelatin certainly precipitates tannin it also precipitates very many other substances including gallic acid ; in fact Jones12 published a list of 88 non-tannins which were precipitated. Trunkel,13 who studied the nature of the reaction, obtained precipitates containing 3 parts of tannin to 1 of gelatin but concluded that since alcohol extracted up to 97 per cent. of the tannin from the precipitates, the process is one of adsorption rather than of chemical combination. Whether chemical or physical however the process has been widely used in combination with the permanganate and other processes.THE HIDE-POWDER METHoD.-The method of estimating tanning capacity by adsorption of tannins with hide powder was first suggested by Bell-Stephens in 1826 and was re-introduced in 1887 by Weiss. Every detail relating to the purification and preparation of the hide powder and to the conditions necessary for obtaining comparable results has been minutely studied and an Officia 296 TEA AND COFFEE THEIR ALKALOIDS AND TANNINS Method has been standardised by the Society of Leather Trades Chemists.l4 Other standard methods are the “International” Method official for Germany, Holland Sweden and Norway,15 and the Official Method of the American Leather Chemists’ Association .16 Whilst the process is of great practical value for leather chemists it has the drawback that it does not always afford aquantitative determination of tannin, since it has been shown that hide powder may adsorb non-tannins (including gallic acid) in addition to tanni11.l~ A striking illustration of this fact is afforded by the recent work of Woodard and Cowland on mat6.ls The application of every recognised test including the gold-beater’s skin test had established the absence of tannin and yet in the hide-powder process 12 per cent.of water-soluble constituents were adsorbed. I t is also interesting to consider the results obtained by this method on a specimen of Chinese gallotannin. This specimen contains 11-5 per cent. of water, 10.7 per cent.of gallic acid about 0.5 per cent. of glucose and (by the colorimetric method) 77.8 per cent. of tannin. Hooper,19 in several determinations by the hide-powder method found 76.8 per cent. of tannin by the cinchonine precipitation method 76.8 per cent. and by the colorimetric method 77.8 per cent. Dr. D. Jordan Lloyd has kindly had the same specimen examined in her laboratory with the following results :-water 11.2; total soluble matter 88.8; non-tans 9-5; substances adsorbed 79-3 per cent. As this tannin has been found independently by Mitchell by Nicholson and by Hooper to contain from 10.7 to 10.9 per cent. of gallic acid it would seem that about 1.4 per cent. of gallic acid was adsorbed by the hide powder in the official method. Adsorfition with Casein.-Nierenstein20 found that casein could be used as a substitute for hide powder and that it did not adsorb gallic acid or glucose.His results however were from 1 to 1.5 per cent. higher than those given by the hide-powder method and he does not appear to have ascertained the nature of the additional adsorbed substances. Spiers21 also used casein as a precipitant in conjunction with the permanganate method the tannin being taken to correspond with the difference between the titration results before and after the precipitation. In this way he obtained concordant results in determining the tannin-content of cider. THE GOLD-BEATER’S SKIN TEsT.-This test is essentially a tanning operation in miniature for it depends upon the fixation of tannin by animal fibre.It was devised by Atkinson and Hazleton,22 and elaborated by Price,23 who showed that it was capable of detecting 0.005 mg. of gallotannin in 1 ml. of water. The value of the method has also been confirmed by several other chemists.24 In applying the test the gold-beater’s skin is first prepared by treatment with very dilute hydrochloric acid then washed tanned for 30 minutes with a dilute solution of the substance under examination washed free from non-tannins and finally stained with a dilute solution of ferrous sulphate. In my experience the method affords the best means yet devised of detecting minute traces of true tannin in the presence of gallic acid and other tannin derivatives MITCHELL A SURVEY OF THE METHODS OF ANALYSING TANNINS 297 THE DICHROMATE TEsT.-The use of potassium dichromate as a reagent for tannins originated with Henry%; it afterwards became known as the Sanio test,26 and was commonly accepted as specific for tannin.I t was shown by Drabble and Nieren~tein,~' however that gallic acid is also precipitated by potassium dichromate, and by Fear,28 that numerous other non-tannin substances react similarly. PRECIPITATION WITH ALKALoIDs.-Although it has long been known that an infusion of cinchona bark would give a precipitate with gallnut tannin it was not until 1834 that P e l o u ~ e ~ ~ suggested the use of quinine as a qualitative test for tannins and in the same year Henry30 asserted that alkaloids as a class were precipitants of tannin. Fear,s1 investigating this commonly accepted belief, found that gallotannin formed precipitates with only six alkaloids (viz.quinine, strychnine brucine cinchonine cinchonidine and caffeine) ; that certain others (e.g. atropine emetine cocaine) gave only a slight turbidity; and that others again, (e.g. pilocarpine aconitine berberine betaine) gave no indications of any reaction. Ware and Smith,32 however have shown that precipitation depends upon the correct adjustment of the pH value and that if the solution is brought to pH 7 to 7-5 by the addition of sodium bicarbonate tannin is precipitated by pilocarpine, emetine cocaine morphine and ephedrine. Wagner,33 having studied the behaviour of strychnine quinine and cinchonine, found that for quantitative work the best results were obtained with cinchonine. In his gravimetric method he precipitated the tannin with cinchonine sulphate, dried the cinchonine tannate a t 120" C.extracted the cinchonine and determined it gravimetrically as sulphate (dried at 120" C.). He also devised a volumetric method in which the tannin was precipitated with a standard solution of cinchonine sulphate in presence of an indicator (rosaniline acetate). His alkaloid solution was standardised by the results of his gravimetric determinations. In 1905 Trotman and Hackford= recommended the use of strychnine 1 mol. of which they found to combine with 1 mol. of tannin; the strychnine tannate was dried first in the air and then in vacuo a t about 60" C. Spiers21 found that the method was accurate for cider tannin but not for gallotannin. (Possibly the explanation is that his " pure " gallotannin contained gallic acid,) Chapmans introduced a refinement into Wagner's method of precipitation with cinchonine sulphate.After drying the cinchonine tannate to constant weight a t 100" C., he determined the nitrogen therein by Kjeldahl's method and from the result calculated the amount of cinchonine in the precipitate thus obtaining a factor by means of which he could calculate the amount of tannin in similar precipitates from infusion of hops. His preliminary experiments were made on a sample of " pure " gallotannin which he assumed to have the formula C14H1,,09.2H20. Next Tatlock and T h ~ m s o n ~ ~ applied the alkaloid method to tea precipitating the tannin with a solution of basic quinine sulphate and drying the precipitate a t 100" C.On the average their precipitates contained 25 per cent. of quinine to 75 per cent. of tannin. Using this method they found Indian teas to contain from 13.3 to 15 per cent. Ceylon teas from 10.1 to 13.9 per cent. and China teas from 7.3 to 10.9 per cent. of tannin. Smith,37 working under the Society's Analytical Investigation Scheme 298 TEA AND COFFEE THEIR ALKALOIDS AND TANSINS studied the application of Chapman’s technique to the determination of tannin in tea.37 After precipitating the cinchonine tea-tannate he extracted the dried precipitate with chloroform to separate caffeine adsorbed by the cinchonine tannate weighed the purified tannate after drying it a t 100” C. and niultiplied the weight by the factor to obtain the tannin present.In this way he obtained results varying from 15.1 to 16-9 for Indian and from 11-6 to 13.5 for China tea. The results were about 1 to 3 per cent. higher than those obtained by the per-manganate method the difference being greater for China than for Indian teas. THE PERMANGANATE METHOD.-For many years the method first devised by Lowenthal,3* of determining tannin by measuring its oxidisability by potassium permanganate was regarded as the standard method and various modifications and simplifications of it were put forward. Thus Monier39 intro-duced the use of an indigo indicator and Procter40 standardised the permanganate on gallic acid instead of on Neubauer’s41 “pure” tannin and used gelatin with salt to precipitate the tannin. The difference between the oxidation values before and after the precipitation was taken to be a measure of the tannins present.The permanganate method gives results for a particular tannin that are comparable among themselves but the oxidation values of the various tannins differ as was shown by Gantter,42 and the permanganate solution therefore requires standardising for each kind of tannin of which the constitution is not known. Moreover as has already been mentioned precipitation with gelatin does not always effect a complete separation of gallic acid from tannin. Hill,43 using Procter’s modification of the permanganate method found China teas to contain from 6.8 to 7-5 per cent.; black teas from 7.8 to 15-0 per cent., and green teas from 9.1 to 24.9 per cent. of tannin. As the tannin was precipitated with gelatin it is probable that some of these figures were too high.IODINE METHODS.-Both gallic acid and tannin absorb iodine and according to Gardner and Hodgsong4 each OH-group requires 1 mol. of iodine. In Jean’s method the reagent is a potassium iodide solution of iodine, standardised on 0.1 per cent. solutions of “pure” tannin and gallic acid and the absorption is determined in alkaline (potassium bicarbonate) solution. In the first titration the whole of the iodine-consuming substances are determined. The tannin is then separated by precipitation with egg albumin and the iodine absorption of the filtrate is determined the difference corresponding with the tannin. Boudet,45 adding an excess of iodine and back-titrating found that 1 g. of iodine was equivalent to 0-469 g.of gallic acid. For mixtures he used hide powder to precipitate the tannin. Cormimboe~f,~~ however found that variable results were obtained either by Jean’s direct or Boudet’s indirect method and that there was no finality in the absorption. To this Jean47 replied that the results are accurate provided that the solution is saturated in the cold with sodium bicarbonate. Colorimetric Method.-The colorimetric method which I devised se\Teral years ago46 is based upon the fact that ferrous tartrate reacts with pyrogallol or the pyrogallic nucleus in gallic acid or tannin to form a violet ink the intensity of the colour of which is proportional to the amount of that nucleus present. It is the MITCHELL A SURVEY OF THE METHODS OF ASALYSIKG TANKISS 299 possible if the constitution of the substance containing the nucleus is known to calculate its amount.This method afforded for the first time an accurate means of determining gallic acid in tannins. The total tinctogenic substances in the tannin are first estimated colorimetrically by comparison with a standard solution of gallic acid (or pyrogallol) the tannin is then precipitated with quinine hydro-chloride and the gallic acid in the filtrate is estimated colorimetrically as before, the difference between the two results corresponding with the tannin in terms of gallic acid or pyrogallol. The accuracy of the method has been repeatedly established (e.g. by Nicholson and R h i ~ ~ d ~ ’ by H~oper,~* and by others) and G l a s ~ t o n e ~ ~ has established the limits for pH for obtaining the maximum colour not only with pyrogallol tannins, but also with catechol tannins.The analytical evidence appears to indicate that “ pure ” commercial tannins are mixtures containing large amounts of gallic acid and this probably accounts for the conflicting and erratic results obtained by various methods standardised on “pure” gallotannin. The specimen of Chinese gallotannin which I used in most of my experiments is probably a mixture of various galloyl glucoses with digallic anhydride for it can be fractionated until it gives a compound which gives a colour closely approximating that which would correspond with a substance of the constitution of Fischer’s penta-digalloyl glucoside. In my original experiments I made a few determinations of the gallic acid and tannin in teas.Not knowing the constitution of tea tannins I had to be content with expressing my results in terms of gallic acid but one advantage of the method is that results previously obtained can be calculated into the pyrogallol equivalent of any formula subsequently established. I found the usual difference between China and Indian teas by this method, a sample of the former containing 3-3 per cent. of tannin (in terms of gallic acid) and one of the latter 7.9 per cent. The respective amounts of gallic acid were 0-84 and 0.80 per cent. The method is not applicable to coffee but I was able to get comparable results by another colorimetric method with osmium tetroxide as the reagent .50 The drawback of this method is that it is difficult to determine when the maximum intensity of colour is reached and that to get concordant results it is necessary to standardise the conditions exactly.REFERENCES 1. 2. H. Trillich and H. Gockel 2. Unters. Nahr. Genussm. 1898 101; Abst. ANALYST, 3. 4. 5. E. Beckmann 2. anal. Chem. 1886 25 527; Abst. ANALYST 1886 11 236. 6 . H. Ruoss 2. anal. Chem. 1902 42 717; Abst. ANALYST 1903 28 116. 7. B. W. Gerland Chem. News 1863 8 54. 8. H. Flick Wagner’s Jahresber. 1860 531. 9. W. H. Krug J . Awer. Chem. SOC. 1895 17 811; Abst. ANALYST 1895 20 283. 10. H. \Yislicenus Z. angew. Chem. 1904 25 801. 1 1 . -4. Seguin Annales de Chim. 1797 20 15. 12. A. E. Jones -ANALYST 1927 52 275. 13. H. Trunkel Biochem. Z. Abst. ANALYST 1910 35 444.G. IVigner PYOC. SOC. Public Analysts 1876 1 235. 1898 23 179. A. Jlanea Chern.-Ztg. Rep. 1905 29 381; Abst. ANALYST 1906 31 80. R. Handtke J . Prakt. Chem. 1860 82 345 300 TEA AND COFFEE THEIR ALKALOIDS AND TANNINS 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. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 60. Cf. Allen’s Comnzercial Organic Analysis 5th Ed. Vol. V p. 137 et seq. Id. p. 153 et seq. Id. p. 157 et seq. F. Jean A n n . chim. anal. 1898 3 145; Abst. ANALYST 1898 23 245. W. A. Woodard and A. N. Cowland ANALYST 1936 60 135. D. Hooper Id. 1925 50 162. M. Nierenstein Chem.-Z&. 1911 35 31; Abst. ANALYST 1911 36 123. C.W. Spiers J . Agric. Sci. 1914 6 77; Abst. ANALYST 1914 39 127. E. Atkinson and E. 0. Hazleton Biochem. J. 1922 16 616. P. H. Price ANALYST 1924 49 25. F. Hardy and F. H. S. Warneford I n d . Eng. Chem. 1925 17 49. 0. Henry J. de Pharm. 1834 20 429. C. Sanio Bot. Ztg. 1863 21 17. E. Drabble and M. Nierenstein Biochevn. J . 1907 2 96. C. Fear ANALYST 1929 54 227. J. Pelouze Ann. Chim. Phys. 1834 57 423. 0. Henry J. de Pharm. 1834 20 429. C. Fear ANALYST 1929 54 316. A. H. Ware and V. Smith Pharm. J. 1933 131 148; Abst. ANALYST 1933 58 703. R. Wagner 2. anal. Chem. 1866 5 1. S. Trotman and J. E. Hackford J . SOC. Chem. Ind. 1905 24 1096; Abst. ANALYST, A C. Chapman J . Inst. Brew. 1907 13 646; 1909 15 360; Abst. ANALYST 1908 33, R. R. Tatlock and K. T. Thomson ANALYST 1910 35 104.H. L. Smith I d . 1913 38 312. J. Lowenthal Z. anal. Chem. 1877 16 33. E. Monier Compt. rend. 1858 46 577. H. R. Procter Chewt. News 1877 36 58; 1878 37 256. C. Neubauer 2. anal. Chem. 1871 10 1. F. Gantter 2. angew. Chem. 1889 577. A. Hill ANALYST 1881 6 95 233. W. H. Gardner and H. H. Hodgson I d . 1909 34 371. M. Boudet Bull. Soc. Chim. 1906 35 760; ANALYST 1906 31 370. C. A. Mitchell ANALYST 1923 48 2. W. N. Nicholson and D. Rhind I d . 1924 49 505. D. Hooper Id. 1925 50 162. S. Glasstone Id. 1925 50 49. C. A. Mitchell Id. 1924 49 162. 1906 31 53. 95; 1909 34 372. 85 ECCLESTON SQUARE LONDON S. W. 1 THE PHARMACOLOGY OF CAFFEINE AND OF TEA AND COFFEE BY G. ROCHE LYNCH O.B.E. M.B. B.S. F.I.C. I will deal briefly with the pharmacology of caffeine and then suggest one or two lines of thought.I understand that it is agreed that tea contains somewhere between 2.5 and 4.5 per cent. of caffeine and coffee 0.5 to 1.5 per cent. THREEFOLD ACTION OF CAFFEINE.-The effects of caffeine on the body can be divided into three groups:-(i) its effect on the central nervous system; (ii) its action on muscular tissue (including heart muscle and that controlling the intestines); (iii) its diuretic action or promotion of the flow of urine. The action of caffeine on the central nervous system is almost entirely a psychical function that is to say it acts on the higher centres of the brain. If it is taken in toxic doses it may exert an influence similar to that of strychnine, namely in producing convulsions.In the course of its action on the central nervous system caffeine facilitates the perception of sensory stimuli and th LYNCH THE PHARMACOLOGY OF CAFFEINE AND OF TEA AND COFFEE 301 association of ideas so that consciousness becomes under its influence more acute. One of the results of that is a condition of wakefulness or increased alertness and so any tendency to drowsiness or fatigue is made to disappear or is less pronounced. A corollary to this is that interpretations of sensory stimuli received by the brain from various external sources become more perfect and accurate. Even more important is the fact that these stimuli are correctly placed in relation to each other. In this respect there is a profound difference between the effect of caffeine and that of cocaine for with the latter in addition to the increased perception of the higher centres enhanced perceptions from the lower centres are also received and the impressions are not so perfect as in the case of caffeine.Thus with ,cocaine the tendency is for the judgment to be impaired; with caffeine the accuracy of the judgment is enhanced. Caffeine also causes a constriction of the musculature of the blood-vessels leading to a rise in blood-pressure and respiration is stimulated. The centres controlling these functions are situated in the lower part of the brain and that is an additional fact in the pharmacology of caffeine. If a person takes a very large dose of caffeine, the process just described is intensified and the result is a confusion of thought and disorders of sensation which are associated with flashes of light in the eyes and noises in the ears-so-called tinnitus.If extreme doses are given this excitation proceeds to restlessness and the receiver becomes tremulous and may develop convulsions such as follow strychnine poisoning. With regard to the action of caffeine on muscle tissue I might remind you that from the medical point of view muscle is divided into three kinds voluntary muscle the working of which is controlled by the will; cardiac muscle a specialised form; and the involuntary muscle such as that in the intestines and the blood-vessels not under the immediate control of the will. Although not definitely known it is believed that caffeine acts directly on the muscle-cells not on the nerve-cells; and the muscular work performed by the person taking caffeine can be increased without the person feeling fatigued in corresponding degree.Here a difficulty arises as it is impossible to say whether or not the abolition of the feeling of fatigue is due to an effect of the drug on the muscles or on the central nervous system. As might be expected from what I have said caffeine is a factor in producing contraction of blood-vessels and intestines and their more vigorous action. There occurs also in those who have taken caffeine a general acceleration of the heart-beat with a diminution of the diastolic period; hence if the dose were large over a period of time the effect on the heart might be definitely un-favourable. In ordinary medicinal doses however the taking of caffeine seems to have no deleterious effect.The cardiac state after large doses of the drug may take the form of auricular fibrillation. Conceivably this might lead to death though actually death from caffeine is rare. With regard to the diuretic action of caffeine the increased flow of urine promoted by it is due to a greater output of water so that the urine itself becomes more dilute than normal; but tested over an appreciable period there is found to be an increase not only in the total urinary output but also in the total solids passed. This elimination of water is among the valuable results of the medicinal use of caffeine as seen in patients who are suffering from dropsy; hence the specia 302 TEA AYD COFFEE THEIR ALKALOIDS AND TANXISS value of the drug in heart failure or in kidney disease.This increased elimination of water has been found t o be due partly to the raised blood-pressure and partly to the specific action of caffeine on the cells of the kidney enabling them to excrete water and to some extent solids too in greater amount. Some of the caffeine is decomposed in the body some is excreted in the urine in an unchanged condition, and some in a partly de-methylated form i.e. as mono- or di-methyl xanthine (caffeine is trimethylxanthine). OVERDOSE.-I have not yet encountered a case in which death was definitely caused by an overdose of caffeine. As much as 60 grains of the drug have been taken at a time and there was recovery from the serious illness. After taking very large doses of caffeine the person manifested,the form of excitation which may be seen in people drunk from alcohol dizziness a ringing and buzzing in the ears trembling confusion of ideas palpitation of the heart and even strychnine-like convulsions.CAFFEINE ADDICTION.-caffeine of course cannot be classed with the drugs which come under the heading of addiction. Those who take caffeine in the form of tea or coffee become accustomed to it and find difficulty in doing without it. Still unlike cocaine and morphine it can be given up without much mental effort or feeling of loss and its indulgence does not cause the serious train of symptoms which follows the habitual taking of cocaine or morphine. In post-mortem examinations I do not believe that any changes occur in those who have drunk largely of the beverages tea and coffee which can be associated with such drinking.I know of no cases warranting the suggestion that either the caffeine or the tannin can produce such effects. CAFFEINE AND SLmPLEssmss.-In conclusion I want just to mention the question of sleeplessness. I am in difficulty over this and it is here that I invite suggestions. I t appears to me very extraordinary that we all know people who will not take coffee as they say they cannot sleep all night after it. Also strong coffee administered per rectum is a common remedy given to patients suffering from any form of narcotic poisoning. Bat if a patient who says he cannot take coffee because it keeps him awake all night is given caffeine citrate in a medicine unknown to him there is often no interference with his sleep.This suggests that the association of tea and coffee with sleeplessness may be largely psychical. Although I have pointed out various attributes of caffeine, such as increased stimulation I feel that there must be some further factor in these beverages which has definite effects as regards sleep but the nature of which can a t present only be conjectured. ST. MARY’S HOSPITAL PADDINGTON W. NORMAN AND HUGHES THE TANNIN-CONTENT OF TEA 303 THE TANNIN-CONTENT OF TEA BY P. J. NORMAN BSc. A.R.C.S. A.I.C. AND E. B. HUGHES DSc. F.I.C. The available methods for the determination of the amount of tannin bodies in tea fail to some extent because of the lack of knowledge of the exact nature of tea tannin and of its oxidation and condensation products.Before describing some work carried out with the object of comparing the results obtained by various methods it may be of interest to remark that tea-tannin a natural constituent of all tea-leaf undergoes some change in the course of the fermentation for the production of black teas-some of it becoming insoluble and some remaining soluble and producing the characteristic colour of the infusion. Green tea little drunk in this country has not undergone this fermentation and the tannin remains soluble and unchanged in colour. Tannin is an important constituent of tea in that it contributes to a considerable extent to those properties which characterise the quality of a tea. There is more tannin in good leaf than in poor leaf and less in stalk than in leaf.A considerable amount of work has been carried out chiefly by workers in the Tea Research and Experimental Stations on the tannin-content of tea-leaf at different stages of growth under different cultural conditions etc. and during the stages of manufacture. Oolong teas are lightly fermented. METHoDS.-The methods that have been used in these comparisons are: (a) Alkaloid precipitation method Smith’s1 application of Chapman’s method for tannin estimation by precipitation from extract or infusion by saturated cinchonine sulphate solution has been used. We use an extract of 1 per cent. w/v diluted after filtration to 24 volumes (i.e. equivalent to 0-4 per cent. w/v) as this does not become cloudy on standing or cooling and moreover it is not necessary to remove caffeine from the liquid before adding the cinchonine sulphate e.g.Tannin extracted (as per cent. of dry tea) r (a) direct from tea extract 15.8 14-8 10-2 11.0 -7 ( b ) after chloroform extraction of the tea extract 16-0 15.0 9.9 11.0 Note that a difference of 0.3 per cent. represents only one mg. of cinchonine tannate. The agreement is within the limits of experimental error. The Lowenthal method,2 by which tannin is determined as the constitu-ent(s) of tea (from infusion or extraction) precipitable by saline gelatin and oxidisable by potassium permanganate with indigo carmine as an indicator of the oxidation. Results are expressed as the gallotannic acid equivalent of the permanganate 1 ml. of 0.1 N oxalic acid = 0.0042 g.of gallotannic acid. (c) The hide-powder method3 the official method of the International Association of Leather Trades’ Chemists which estimates tannin as the total solids removed by freshly-chromed hide-powder from an extract or infusion. (b 304 TEA AND COFFEE THEIR ALKALOIDS AND TANNINS RESULTS.-AL~ results given in the paper are as percentage of dry tea. Table I gives amounts of tea-tannin as obtained by the different methods in typical samples of unblended teas. TABLE I Extractable tea-tannin (per cent. of dry tea) A. Black Teas Lapsang Souchong (China) tea Darjeeling Orange Pekoe . . Ihemun (China) tea . . . . Ceylon Orange Pekoe . . . . Java Orange Pekoe . . . . i’,nnam Orange Pekoe . . . . Annam Souchong . . . . Japan black t e a .. . . . . Syasaland Broken Pekoe . . Mcthod (a) 10.1 10.4 14.7 12.8 14.0 12.9 14.7 13.4 10.8 Method ( b ) Method (1.) 6.2 10.1 5.3 -12.0 14.4 11.1 11.6 9.6 13.3 0.2 11.1 13-5 -7.4 -6.7 9.6 B. Green apzd OoLong Tens Noyune Young Hyson (green tea) 12.7 13.5 14.2 11.9 Formosa (Oolong) tea . . . . 15.1 15.4 17.9 I\ 1 o y u TI e G u n p ow d e r ( g r e e 11 tea) 9.5 10.1 These results are represented on Graph I. I t will be seen that in Group A the results by method ( b ) the Lowenthal method are always lower sometimes considerably so than those obtained by cither the cinchonine method ( a ) or the hide-powder method (c) and that the results by these last two do not differ greatly. These teas are all black teas-i.e. teas which have undergone full fermentation in manufacture.The teas of Group B do not show this lower Lowenthal result. These are two green teas which have not been fermented and an Oolong tea (lightly fermented). These results suggest that the fermentation has affected the tea-tannin in such a way that the permanganate required for its oxidation has decreased; this decrease is greatest (as a proportion of the cinchonine-precipitated tannin) in the China tea which is more fully withered and fermented than the usual Indian or Ceylon black teas. The amount of cinchonine tannate precipitate is apparently not so affected. It may be significant that we have found that the tannins of cacao can be subjected t o severe oxidation treatment without appreciably altering the amount of cinchonine precipitate given.I n Table I1 and Graph I1 we give similar results for commercial blends of tea. L411 with the exception of Nos. 1 and 2 (blends sold simply as “tea”) are teas for which some specific claim such as “Digestive,” “Invalid,” etc. is made. With regard to the others, which are black teas of the Indian or Ceylon (mainly Ceylon) type or blends they are seen to be very similar in tea-tannin content by whichever method it is estimated . Teas Nos. 1 2 4 5 6 are black teas similar to those of Group A Table I , No. 5 is a China tea obviously similar to No. 2 NORMAN AND HUGHES THE TANNIN-CONTENT OF TEA 305 G W H 1 REILBTIVE TAMIN CONTENTS OF UNBUNDED TEAS. TRBLE IA BLRCK UNUNDED TEAS MOYV-h!E ~ M~ YOUNG HYSON 0 HO# H0 Y / MOYUM ,@GUNPOWDER 0 0 0 0 LOWENT.METHOD. LOWENTILKC METHOD. TABLE 91 Ten. 1. (Blend of Indian and Ceylon teas). . 2. (Blend of China teas) . . . . “Special” teas . . . . . . Extractable tannin (as per cent. of dry tea) blethod ( a ) Method ( b ) Method (c) f A \ 15-8 11-0 16-5 17.3 11.8 16.4 15.4 16.6 14.5 14.0 7.4 14.6 13-8 8.6 12.9 13.2 13.2 14.2 14-0 9.4 13.6 14.1 10-4 14.2 12.2 12.6 13-306 TEA AND COFFEE THEIR ALKALOIDS AND TANNINS having lower results for the Lowenthal method than for the cinchonine or hide-powder method. The other samples although they are black teas and accordingly have lower Lowenthal than cinchonine values show still lower results for the hide-powder method.This may possibly be due to the grading of the tea; we have some evidence suggesting that fannings,”” of which these teas mainly consist behave in this way. R E L J T ~ T m m CONTENTS OF BLENDED TEAS. TEA CINCHONINE H m - PO WDER H . O D . I METHOD. Z 8 - m METHOB. We have in general used only the cinchonine method for the determination of tea-tannin and some further results by this method for such teas as those in Table I1 are given in Table 111. * The broken tips etc. of the rolled leaf broken off during manufacture and separated by sieving NORMAN AND HUGHES THE TANNIN-CONTENT OF TEA 307 TABLE I11 Extractable tannin (per cent. of dry tea) Method ( a ) Indian and Ceylon blends . . 3. 4. 14.6 ii } “Special” teas . . 8. 16-2 Nos.1 to 4 are ordinary commercial blends of Indian and Ceylon black teas, and Nos. 5 to 8 are teas sold as “Digestive,” etc. teas. TABLE IV TANNIN-CONTENT OF UNBLENDED TEAS Extractable tannin (per cent. of dry tea) Method (a) China green teas Moyune Gunpowder . . 9.5 Moyune Young Hyson . . 12.7 China black teas Lapsang Souchong 10.4 Keemun . . . . . . 10.1 N. India Patrakola broken Pekoe . . 14-1 Chalouni broken Orange Pekoe . . 16-8 Darjeeling Orange Pekoe . 14.7 Assam broken Pekoe . . . . 14-3 S. India Letchmi broken Pekoe . . . . 15-6 Travancore . . . . 17.2 Ceylon Broken Pekoe . . . . . . 13-4 Broken Orange Pekoe . . . . 15-5 Darjceling Pekoe . . . . . . 13.4 Assam Orange Pekoe . . . . 15.1 Orange Pekoe . . . . . . 12-23 Nyasaland broken Pekoe .. . . 12.9 Java Orange Pekoe . . . . . . . . 14.0 Java broken Pekoe . . . . . . 13.6 Indo-China Annam Souchong . . . . . . 13.4 Annam Pekoe . . . . . . 14.6 Annam Orange Pekoe . . . . 14-7 Formosa Oolong tea . . . . . . . . 15-2 Japan black tea . . . . . . . . . . 10.8 Korea broken Pekoe . . . . . . . 16.6 In this table are given the tannin-contents (by the cinchonine method) of a more extended range of unblended teas than are recorded in Table I. It is however not actually the whole amount of tannin that can be extracted by prolonged boiling from the tea which concerns the user but the amount dissolve 308 TEA AND COFFEE THEIR ALKALOIDS AND TANNINS out in the ordinary way of making tea. For such tests we make a standard infusion by adding 29 parts by weight of boiling water to one part of tea and decanting (and filtering) after 3+ minutes.The factors which obviously would be expected to influence the amount of tannin obtained in an infusion of tea are:-(i) the length of time of the infusion; (ii) the ratio of the amount of tea to water; (iii) the temperature of the water. (i) Time of Infusion.-A sample of a commercial blend of Ceylon and Indian tea of medium price gave the following results for infusions made during various periods of time : TABLE V STANDARD INFUSION VARYING TIME Time of infusion in minutes 2 3 4 5 6 7 8 9 10 Present in infusion (per cent. of dry tea) (Method (a)) solids 11.5 19.4 12.5 20.6 12-5 22-5 13.3 21-4 14.2 21.2 14.1 21.7 13.7 22.7 14-2 21.8 13.8 21.9 Tea-tannin -Il From these figures it is seen that there is an increase in the amount of tannin up to 6 minutes but that the non-tannin soluble solids go more quickly into solution.(ii) Ratio of Tea to Water.-A similar type of commercial black tea gave the following results for increasing ratio of tea to water. Conditions as for Table V and time of infusion 3& minutes. TABLE VA Tea:Water (w/v) 0.2 29.8 0.4 29.6 0-6 29-4 0.8 29.2 1 29-0 2 28 3 27 Amount in infusion (per cent. of dry tea) Tannin Non-tannin solids c A \ 11.0 11.0 11.0 10.9 10.9 8-8 9.0 21.0 21.8 20-9 21.5 23.2 19.9 20.6 These figures indicate that the amount of tea does not cause appreciable decrease of the degree of extraction of tannin until the proportion of tea to water is greater than 1 to 29.(iii) Temperature of the Water.-Table VB gives results for infusions prepared with water a t various temperatures otherwise with " standard " proportions and time NORhlAN AND HUGHES THE TANNIN-CONTENT OF TE.4 309 TABLE VB 3 i mins. 1 in 30 Amount in infusion (per cent. of dry tea) Temp. of infusion I A > "C. Tannin Non-tannin solids 60 5.2 14.4 70 6-9 18.0 80 0.7 19.0 90 11.5 20.5 100 12.5 21.1 Clearly the temperature of the water is a matter of considerable importance. Under the standard conditions which we employ the temperature of the liquid during the infusion of 3+ minutes does not fall below 90" C. (about 92" C. a t the end of the infusion). Table VI shows the amount of tannin removed by infusion under "standard" conditions from the same teas as those for which total extractable tannin-contents were given in Table IT.TABLE VI STANDARD INFUSIONS Tea 1. 2. 3. 4. 6. 7 . 8. 9. L a. Tannin extracted (per cent. of dry tea) 9-6 4.8 10.4 10.1 5.2 10.5 9.2 10.4 9-9 Method ( a ) I t will be noticed that as for the total tea-tannin the figures are much the same for all the non-China black teas and likewise for the two China teas and that they are in general in the same relative order as for the total tannin of the same teas though the actual amounts removed are less (about 50 per cent. for China teas Nos. 2 and 5 and about 70 per cent. for the other black teas). The results we have given in the paper indicate the importance of specifying exactly the method used (and also the procedure) for estimation of tea-tannin, though for comparisons of teas of the same type (China non-China black teas, green teas) the relative results are not seriously affected by the choice of method.We desire to thank J. Lyons & Co. Ltd. in whose laboratory this work was carried out for permission to publish. REFERENCES 1. H. L. Smith ANALYST 1913 38 312. 2. 3. J . Lowenthal 2. anal. Chem. 1877 16 33. Allen's Commercial Organic Analysis 5th Ed. Vol. V p. 137 et seq 310 TEA AND COFFEE THEIR ALKALOIDS AND TANNINS “TANNINLESS” TEA BY H. H. BAGNALL B . S c . F.I.C. During the past six years I have received at intervals a number of samples of packet teas bearing on the labels statements concerning the tannin-content.In nearly every instance this statement indicated either that no tannin at all was present (which is an obvious mis-statement) or that the amount of tannin was less than was usual in teas of other types. As a rule the inference was drawn that the tea would be “more digestible,” would ‘‘ promote digestion ” or as one bold statement asserted would actually “ cure indigestion.” These statements appear to be simply assertions made without any particular basis of scientific fact. The tannin was determined in all these samples and in a few others for com-parison by Thomson and Tatlock’s quinine method,l the directions given being followed exactly. Smith’s cinchonine method2 was tried in a few cases but the process was longer and troublesome emulsions were sometimes obtained during the extraction of caffeine by chloroform.However the figures obtained on the same tea by the two processes did not differ significantly being usually a little lower by Smith’s method. As the figures obtained were intended to be comparative only between one tea and another and in view also of the fact that all methods are only ap-proximate the exact nature of the tannin being unknown the simpler quinine method was followed throughout the series of determinations. The amount of tannin varied for Indian and Ceylon teas from 9.9 to 16.4 per cent. the average being 12.8 per cent. Two China teas which were examined each contained 8-6 per cent. of tannin. Of the 31 samples in which the tannin was determined 16 were made the subject of enquiry, and although no case was taken to Court the packers in every instance recognised the weakness of their position and agreed to omit references to tannin which were regarded as contravening the provisions of Section 30 of the Food and Drugs (Adulteration) Act.In the following table is given a list of the fifteen samples regarding which action was taken the offending portion of the label being quoted: Twenty-two samples of 29 fell within the range 11.0 to 14.1 per cent. Date Tannin Per Cent. Extracts from labels Aug. 1929 12.5 “Free from crude tannin found in all ordinary tea, therefore good for indigestion gastritis etc. . . . Cures indigestion.” “ Contains the maximum of theine with the minimum of tannin,” and other statements implying that its value was greater than that of other teas.Apr. 1930 13.9 “ Remarkably free from objectionable tannic acid, the chief cause of indigestion to users of ordinary full leaf teas. Contains only the delicate harmless portions of the leaf avoiding the coarser parts which contain injurious ‘tannics’.” Feb. 1930 14-9 (caffeine 3-36 311 Date May 1930 May 1930 June 1932 Jan. 1933 Jan. 1933 Mar. 1933 May 1033 July 1933 Jan. 1934 Jan. 1934 June 1934 June 1935 BAGNALL “ TANNINLESS ” TEA Tannin Extracts from labels Per Cent. 14.0 “ Free from crude tannin. Practically tanninless.” 12.7 “Fine tea contains very little tannin and conse-quently . . . this tea can be used freely . . .by persons who suffer from indigestion etc.” 14.7 “No crude tannin present.” 12-7 “ Free from tannin.” 14.9 “ Tannin minimised.” 11.2 13.6 14.1 13-9 “ Practically free from tannin.” “All stalks wherein lies the tannin eliminated.” “Composed only of the tips of leaves and therefore, “ Contains all the essential goodness without any t anninless.’ ’ injurious tannin. ” 8.6 (China) “ Practically free from tannin.” 11.7 “ Contains the minimum of tannin.” 16.4 “Contains the maximum of caffeine and the mini-mum of tannin. Digestive because non-tannic. Free from stalks etc. which contain crude tannin.” All the teas mentioned in the table were so-called “Digestive” teas and were finely ground to give the appearance of the popular “Leaf tips.” I n addition to the samples contained in the table one or two others were labelled in a dubious manner.The label of one (June 1934) stated that one spoonful would more than equal two spoonfuls of ordinary leaf. The implication here of course was that the tea would go twice as far but the actual fact was that owing to the tea being finely ground the spoon would hold a greater weight than it would of a coarser variety. The water extract determined on a 0-25 per cent. solution was 39-9 per cent.-an average figure. The label of another sample (June 1934) contained the statement that a half-pound would go as far as one pou5d of ordinary tea and the tea was described on another part of the packet as “double strength.” The water extract in this case was 42 per cent. and the claim made was obviously preposterous.A sample (January 1934) was also stated to go twice as far as ordinary tea, and was further described as a great nerve tonic being composed of “tea tips of immense strength.” Another interesting sample (November 1933) was labelled “ Rich in vitamins,” and was described as “a blend of Empire leaf combined with the tiny leaves of a wonderful tropical plant which has remarkable curative properties in cases of indigestion rheumatism neuritis etc.” Black tea contains no vitamins except a possible trace of vitamin E . The tropical plant referred to was mat6 which was present to the extent of about 7 per cent. A sample (May 1933) contained 14.0 per cent. of tannin and an analysis made by “ a celebrated London analyst,” appeared on the label giving a figure of 11.2 per cent.I t was not stated however by what method or at what dat 312 TEA AND COFFEE THEIR ALKALOIDS AND TANNINS this analysis had been carried out. It was described as “ Real edge and leaf tip tea,” and claimed that it could be “enjoyed by persons of weak digestion owing to its low tannin-content as compared with that of common coarse leaf tea.” Representations to the firms concerned in the packing of the above samples have resulted in most cases in a modification of the statements to which objection was taken. I t is believed that by reason of the administrative action taken there are now very few teas on the market to the labels of which serious objection can be taken and incidentally there is good reason for thinking that the persuasive methods employed to induce the packers to revise their labels were far more successful than the more forcible (and expensive) method of taking legal proceedings.REFERENCES 1. R. K. Tatlock and K. T. Thomson ANALYST 1910 35 104. 2. H. L. Smith id. 1913 38. 312. CITY ANALYST’S DEPARTMENT BIRMINGHAM DISCUSSION Mr. D. M. FREELAND said that he had heard that tea-blenders made different blends of tea for different parts of the country because of the variations in water. Did it really make any difference whether distilled or tap water were used for the infusion ? Dr. H. H. MANN replied that the character of the water undoubtedly had an influence and that this was taken into account in preparing blends for use in different places. Dr. H.E. Cox observed that according to recent papers in the Zeitschrift Untersuchungs Lebensmittel there was still much doubt as to the presence of chlorogenic acid in coffee. Some said there was none; others showed about 7 per cent. Discrepancies were also apparent in relation to caffeine. I t was known that chlorogenic acid was easily hydrolysed but he would like to know what was the relationship between caffetannin and chlorogenic acid and whether there was any probable explanation for the manifest discrepancies in the published papers. Dr. LAMPITT remarked that the papers had demonstrated the very indefinite state of our knowledge and it seemed to him that as Dr. Nierenstein had isolated some beautiful crystalline products it would be very valuable if they could be submitted to various standard methods of analysis in order to see if some corre-lation could be obtained.Thus if a definite tannin were obtainable from tea, it should be possible to get some factor to which they would be able to refer their results for tea-tannins. He really did not know. The whole chemistry of caffetannin was most unsatisfactory. There were three possibilities-either that the acid was the caffetannin or that the acid was the parent of the caffetannin or that it was derived from the caffe-tannin. He agreed that if practicable the suggestion of Dr. Lampitt was a valuable one but until the crystalline derivative could be obtained in quantity it would not be possible to co-ordinate results. He had not yet determined the colorimetric ratio of the tannin.Mr. R. F. INNES said that he regarded these papers as valuable contributions t o the subject especially the survey of the methods of analysis. Apparently Dr. Mitchell did not rate the hide-powder method very highly but it was accepted in the leather trade as an empirical method which gave results approximating closely to those obtained in actual tanning practice. The colorimetric method Dr. NIERENSTEIN referred Dr. Cox to a chapter in his book DISCUSSION 3 13 d3vised by Dr. Mitchell gave an accurate measure of the tannin in solution from one point of view but tannin was a variable substance and it did not seem likely that the method would distinguish between pyrogallol and catechol tannins. Presumably the method gave the results for both in terms of one.Mr. A. W. KNAPP asked whether the colorimetric method was very sensitive t a pH. Some years ago there were two papers (one by Adams and one by Jensen) on the determination of tannins in cocoa. In each the tannin was determined by the cinchonine method. Apparently the only difference was that Adams took the bean as it was whereas Jensen did what he called “neutralising the natural acid of the bean” before determining the tannin. Would a slight difference in pH make a difference? Dr. MITCHELL replying to Mr. Innes said that it was not possible to dis-tinguish colorimetrically between pyrogallol and catechol tannins in admixture, unless possibly by rigid control of the $H conditions. The $H also had a pronounced effect upon the precipitation of tannin by cinchonine.Mr. H. S. REDCROVE pointed out that a great deal of the coffee made in England did not taste like coffee. Thus it would seem that the English public was more interested in the caffeine than in the flavour. He thought this question of flavour was important. It played a large part in the price which could be got for tea in this country. From the point of view of flavour it did not matter what tannins one had there beyond the fact that tannin gave a certain astringent taste; the flavour was due to an essential oil. Efforts had been made in certain perfume factories in France to extract the flavour from tea. He wanted to stress the point that taste and flavour were different phenomena because it had struck him that they might go away from this meeting without anything having been said to the effect that the flavour could not possibly be due to a non-volatile substance such as tannin but must be due to essential oil.Dr. H. H. MANN said that his long connection with the tea industry had led him to think that the present meeting would interest him and it had certainly done so. In the first place as a representative of the producers the question of the relation of the various constituents to the value of the tea was to him the most important matter before the meeting. ,411 investigation up to the present seemed to show that the proportion of caffeine in the tea did not affect its market value at all. One could get low-priced tea with high caffeine-content and high-priced tea with a small proportion of caffeine. Regarding the relationship of the amount of tannin to the value of tea it might be said that generally speaking, the higher the amount of tannin the higher the price of the tea.Most people thought that the opposite was the case and indeed such a relationship was not by any means universal but the statement represented the general position when comparing teas of a similar class. The exceptions were sufficiently numerous, however to make it important for further study. His own idea was that the reason for the absence of a constant connection between the amount of tannin and the value was that the tannin contained in a commercial sample of tea was not a single body and was not the tannin originally contained in the leaf, but was a mixture of tannin and tannin dcrivatives in an infinity of stages and conditions of oxidation.In all the analyses that were normally made all these were taken together and while this was done it was quite impossible for the analyses to show any connection with the market value. By varying the con-ditions of manufacture the proportion of these oxidation products could be varied very greatly and so the value of the tea produced could be very greatly modified. Until the analyses employed could differentiate between the various forms in which the tannin and tannin products occurred in commercial tea it was probable that determinations of tannin would be absolutely useless in connection with the valuation of the tea. At one time he thought he had got a method which would serve for this purpose but while it applied in some cases it gave results in other 3 14 TEA AND COFFEE THEIR ALKALOIDS AND TANNIKS which were contrary to the opinion of the market and the price of the tea.A t present he doubted whether the determination of the tannin by any known method was of any use to the consumer and they were compelled to go back to the taster for the value of tea. Until analysts could go further than they could a t present, the determination was not worth the time spent on it. I t must be remcmbered that the tannin of tea was of a different type from most of the tannins that were usually discussed. The tannin of tea was not ;t residual product. I t was present in largest quantity in the earliest leaf growth, and it seemed probable that in tea tannin took the place of starch in connection with the metabolism of the plants.He was extremely interested by the formulae presented by Dr. Nierenstein as representing the composition of the tannin of tea-leaf but there seemed a good deal of probability as a result of the work of Shaw and Jones in South India that there were differences in the structure of tannin in different types of tea and even between that in tea-leaf a t different times of the year and that such slight differences were very important in connection with the quality of tea. This matter was however in such an early stage of investigation that hc would not pursue the subject at the moment. At present he was very anxious to emphasise the point that he did not think that analyses of tea merely giving a total figure for tannin were of any value whatever.What was needed now was not the determination of tannin as a whole, but some method of fractionation of the tannin if such analyses were ever to be of use in determination of the value of teas. It seemed to him a waste of time and money to detcrmine the tannin when it bore no certain relationship to the quality and flavour of the tea. He could illustrate the importance of the tannin in tea. I n Germany to-day there was a process (which he had fully investigated) whereby tannin could be taken from tea; and having taken it out what had one left? Nothing for it did not taste remotely like tea. With regard to tannin in tea he was perfectly certain that legitimate traders in tea did not want to be concerned in any way with the alleged tannin question. Mr. Bagnall’s paper showed quite definitely that there was a number of firms who were stimulating sales of tea by pandering to the few. He was glad to hear DI-. Roche Lynch say that in no case of a post-mortem examination of tea-drinkers had he found anything abnormal attributable to the drinking of tea. That was one of the most important statements that had been made. Dr. HUGHES stated that they had carried out a few tests on the application of Mitchell’s colorimetric method to the determination of tannin in tea infusions. In order to obtain a result for the tannin equal to that given by the cinchonine precipitate method it was found that the pyrogallol equivalent as determined colorimetrically should be multiplied not by the factor 2 but by a factor of about 1.5 for green tea 2.1 to 2.2 for non-China tea and 2.4 to 2.5 for China tea. This indicated that the fermentation of tea-tannin affected the colour equivalent in this method in the same way as the permanganate absorption in the LowenthaI method the cinchonine precipitate remaining much the same. The gallic acid figures obtained were 1.25 per cent. for green tea 1.3 per cent. for non-China black tea and 0.9 per cent. for China tea. In estimating the gallic acid in the filtrate from the cinchonine precipitate it was found necessary to precipitate the cinchonine with the requisite amount of sodium carbonate (without excess). Dr. Hughes also drew attention to the question of the ratio of caffeine to tannin in tea. Harler has pointed out that there was no constant ratio and that it might vary from 1 3 to 1:12 according to the strength of the infusion and they had found this to be so the ratios varying even for unblended teas from 1 1.7 to 1 4.5 and also varying with the time of the infusion. Dr. LAMPITT entirely agreed with Dr. Pulann. I t was for the flavour that people took it

ISSN:0003-2654    

DOI:10.1039/AN9366100288

出版商:RSC    

年代:1936

数据来源: RSC

摘要:

BERRY FVKTHER E:SPEKIMENTS LVITH =IDSORPTION INDICATORS 3 1 3 Further Experiments with Phenosafranine, Tartrazine and Rose Bengal as Adsorption Indicators BY A. J. BERRY M.A. IT is interesting to review the modifications that have been introduced into the original volumetric process for the reciprocal determination of halogens and silver devised by Gay Lussac just over one hundred years ago. Various refinements, chiefly due to Stas have resulted in the method becoming one of the most accurate known to chemists. Rapidity of working at the expense of-a certain degree of accuracy was realised by Nohr by employing potassium chromate as an indicator for titrating chlorides in neutral solution. Mohr’s method was followed by Volhard’s well-known thiocyanate method for determining silver in acid solution.In more recent years argentometric methods have been further improved by the use of adsorption indicators due to Fajans. I t is no exaggeration to claim that the introduction of adsorption indicators is the most important advance which has been made in this branch of analysis since the time of Mohr and of Volhard, a:id the value of these indicators may be judged by their rapidly increasing use. Recently Fajans has published a most interesting monograph on this subject, entitled “Adsorptionsindikatoren fur Fallungstitrationen,” in which the theory and practical applications of the subject are discussed in detail together with a useful index of the 1itei-ature.l In the present paper some new experiments are described which amplify results already p u b l i ~ h e d ~ ? ~ and also illustrate the use of these indicators in various types of volumetric determinations which are commonly effected by other methods.Two of the indicators namely phenosafranine and tartrazine give satisfactory results for the titration in nitric acid solution up to an acid concentration of about normal. I t is therefore possible to employ either indicator in conjunction with potassium bromide as the titrant for any determination which would otherwise be effected by Volhard’s method. The analysis of arsenates in approximately *I’/lO concentration and of silver in Levol’s alloy in solutions of N/100 concentra-tion may be quoted by way of illustration. The third indicator Rose Bengal, was found by Fajans and \V01ff4 to be useful in the determination of iodide in presence of chloride by titration with silver nitrate but its use is restricted to neutral or very feebly acid solutions.The use of these indicators in the analysis oi‘ mixtures of cyanides chlorides and iodides is illustrated. Finally a method fc .I- determining halogens in electrolytes of limited ionisation involving the use of tartrazine is described. In carrying out titrations in which adsorption indicators are used for deter-mining the end-point it is always desirable to adjust conditions to facilitate flocculation of the silver halide from the colloidal condition. As has been noted previ~usly,~ flocculatioii may be effected by adding a bivalent electrolyte such as strontium nitrate. In the absence of substances which give rise to complication 3 16 BERRY FUKTHEK EXPERIMEKTS \VITH PHESOSAFRANINE, this should always be done.\Then however such a proceeding would involve the production of a sparingly soluble precipitate it should be avoided. No precipitate other than the silver halide should be present when an adsorption indicator is used. For this reason in one of the methods described below for the analysis of mixtures of cyanide chloride and iodide as potassium bitartrate is used for eliminating the cyanide from solution strontium nitrate must not be added since it would involve the separation of strontium tartrate. In these and in other similar cases flocculation must be effected by patient and sometimes prolonged shaking. 1. MIXTVRES 01; CYASIDE CHLOKIDE AKL) IouIuE.-In the titration of potassium cyanide in such mixtures with silver nitrate the end-point of the reaction corresponding with the complete production of potassium argenticyanide can be seen perfectly well at the slightest appearance of permanent opalescence, particularly if a black surface is placed under the titration vessel; and so far as my experiments are concerned there is no advantage in having an adsorption indicator present.Even at so low a concentration as A'i50 accurate end-points are obtained without difficulty. If however the further titration is attempted in presence of such an indicator the silver cyanide adsorbs the dyestuff with gradual change of colour before the end-point is reached. I t was found to be absolutely essential to eliminate the cyanide before proceeding to the determination of the halide with the aid of an adsorption indicator in the analysis of mixtures.Special experiments with mixtures of potassium cyanide and potassium bromide (approximately iV/lO) showed that the two constituents could be determined with accuracy by first titrating the cyanide with silver nitrate directly then taking a measured volume of the original solution eliminating the cyanide by boiling with a small quantity of nitric acid and titrating the bromide with silver nitrate with phenosafranine as indicator. Alternatively after determination of the cyanide, a known excess of silver nitrate and a little nitric acid are added the silver cyanide and bromide are removed and the silver remaining in solution is titrated with a solution of potassium bromide either phenosafranine or tartrazine being used as indicator.Complete agreement was realised between the two methods and satisfactory results were obtained at a concentration of N/50. Fajans and Wolff4 have shown that it is possible to determine a chloride and an iodide together in the same solution by titration with silver nitrate with the use of two different adsorption indicators the success of the method dep$nding partly upon differences in adsorbing capacity of the two anions and partly upon the relative degrees of insolubility of the two silver halides. In the presence of 1 arious halogenated fluoresceins such as Rose Bengal (dichloro-tetraiodofluores-cein) a marked colour change takes place when silver iodide is precipitated com-pletely and the chloride remaining in solution can be determined by titration with silver nitrate with fluorescein as indicator.Numerous experiments on various mixtures of chloride and iodide have verified the findings of Fajans and Wolff so far as the accurate determination of the iodide with the aid of Rose Bengal is concerned. However the titration of the chloride remaining in solution with silver nitrate and fluorescein as indicator, was found to be altogether unreliable. Very satisfactory results were nevertheles TARTRAZINE AND ROSE BENGAL AS ADSORPTION INDICATORS 317 obtained by decanting the liquid through a filter washing the precipitate with very dilute nitric acid and titrating the filtrate with silver nitrate with pheno-safranine as indicator. Numerous experiments also showed that the colour-change with Rose Bengal coincides strictly with the quantitative precipitation of silver iodide and without any co-precipitation of silver chloride.Moreover it was found that whilst Rose Bengal cannot be used in the presence of strong acids, this indicator gives excellent results in presence of very weak acids. In the analysis of cyanide-chloride-iodide mixtures it was found convenient to effect elimination of the cyanide by boiling the solutions with a small quantity of potassium bitartrate for about a quarter of an hour. Iodide and chloride could then be determined in the resulting (cooled) liquid in the manner indicated. Three separate titrations are thus required to determine the constituents of a cyanide-chloride-iodide mixture.First a portion of the solution is titrated directly with silver nitrate without an indicator to the opalescent stage for deter-mination of the cyanide. Secondly a fresh quantity of the solution is boiled with a slight excess of potassium bitartrate to eliminate the cyanide and the iodide and chloride are determined as described above. In the earlier experiments the chloride was determined by difference as follows :-Excess of silver nitrate, followed by a little dilute nitric acid was added to a fresh quantity of the solution, the mixed precipitate of silver halides and cyanide were removed by filtration and the silver remaining in solution was titrated with a solution of potassium bromide, tartrazine being used as indicator. The accuracy of these methods was verified by experiments on a large number of solutions containing the constituents in varying proportions.Iil the first place, some titrations of a mixture of potassium cyanide and bromide (approximately N/10) may be quoted to illustrate the agreement between the titration values for the potassium bromide (i) after removing the cyanide by boiling with a little normal nitric acid and (ii) by adding excess of silver nitrate removing the silver cyanide and bromide and titrating back with potassium bromide Volume of silver nitrate required for visible opalescence (i) 14.7 ml. lrolume of silver nitrate required for visible opalescence (ii) 14.7 ml. Volume of silver nitrate required after boiling out the hydrocyanic acid 31.8 ml. (indicator Volume of silver nitrate calculated from the potassium bromide back-titration 31-8 ml.(indicator tar trazine) phenosafranine) When these solutions were diluted to one-fifth of their original concentrations the same titration values were obtained. Experiments on the same lines on various cyanide-chloride-iodide mixtures showed satisfactory agreement in the titration values both for the cyanide and for the iodide and chloride. One example may be quoted in which the chloride was determined by the “difference” method, after determination of the cyanide and iodide by back-titration with potassium bromide and tartrazine 318 BERRY FIJRTHER EXPERIMENTS WITH PHENOSAFRANINE, For cyanide For iodide For chloride Observed silver nitrate titrations . . 12.2 ml.10.6 ml. 10-6 ml.* Calculated , 7 9 . . 12.2 , 10.6 , 10.45 ,, * From back-titration with potassium bromide. Two more examples may be quoted. In these the iodide was titrated in the usual way in presence of Rose Bengal and the chloride was determined in the filtrate from the silver iodide in very dilute nitric acid solution phenosafranine beicg used as indicator. For iodide For chloride (a) Observed silver nitrate titrations . . . . 40.8ml. 16.3 ml. Calculated , J J ) ) . . . . 40.7 ) 16.45 ,, (b) Observed , 7 , . . . . 10.2 , 40-65 ,, Calculated , ? , . . . . 10.2 , 40-5 ,, 2. COMPARISON OF RESULTS OBTAINED IN DETERMINATIONS OF SILVER IN ACID SOLUTION BY VOLHARD’S METHOD AND BY TITRATION WITH POTASSIUM showed most satisfactory agreement. For work at concentrations of about N/10 either adsorption indicator is equally useful but at much greater dilution phenosafranine is preferable.By way of illustration the following analyses of Levol’s alloy-* in N/100 concentration may be quoted: Pure silver (0.4424 g.) was dissolved in nitric acid and the solution was diluted t o 500ml. Of the alloy 0.4736 g. was dissolved and the solution was diluted to 500ml. Quantities of 50ml. were taken for each titration. BROMIDE WITH PHENOSAFRANINE OR TARTRAZINE AS INDICATOR.-The results Volumes of potassium bromide using Volumes of potassium phenosafranine as thiocyanate required indicator required For the pure silver solution . . 39.7 ml. 40.95 ml. For the alloy solution . . . . 30.35 , 31-2 ,, The calculated percentages of silver in the alloy are 71.4 by Volhard’s method and 71.2 by the adsorption indicator titration method.The two methods were also compared in the well-known silver method for the determination of arsenates. Solutions of approximately K/lO concentration were used in these experiments. The silver arsenate was precipitated by adding an excess of silver nitrate in presence of a little nitric acid and a large excess of sodium acetate to measured volumes of a solution of sodium arsenate. The washed precipitates were dissolved in N nitric acid and the solutions were titrated with potassium bromide with tartrazine as indicator and by Volhard’s method. The weights of silver found per 20 ml. of the solution of sodium arsenate were 0.330g. by Volhard’s method and 0-331 g.by the adsorption indicator method. 3. TITRATION OF HALOGENS IN THALLOUS-THALLIC SALTS.-It has been shown that chlorine in thallic chloride cannot be determined by Volhard’s method * Levol’s alloy discovered in 1854 by Levol is an alloy of silver (71.9 per cent.) and copper. The concentration N/100 refers of course to the concentration of silver in the solutions of alloy in nitric acid which were being titrated TAHTRAZINE AND ROSE BENGAL AS ADSORPTION INDICATORS 319 on account of the oxidising action of thallic ions on thiocyanate (Cushman,s Berry6). Further as the thallic halides are weak electrolytes their behaviour on titration with silver nitrate with the use of adsorption indicators is irregular (Berry3). It was found however that the method of reduction with zinc amalgam in presence of a little dilute sulphuric acid applied to the determination of the total chlorine in chloropentammine cobaltic chloride could be used for the determination of halogens in these compounds.A 2 per cent. zinc amalgam was used for the reductions and the reduced solutions were run from the burette into a known quantity of a silver nitrate solution tartrazine being used as indicator. Thus, 3.074 g. of thallous thallic chloride (thallium sesquichloride) was reduced and the solution was diluted to 200 ml. A solution of 0.3557 g. of silver in nitric acid required 36.8 ml. of the reduced solution corresponding with a total weight of thalliuni sesquichloride of 3.073 g. Again 1.768 g. of thallous thallic bromide (thallium dibromide) was reduced and the solution diluted to 200 nil.Twenty ml. of silver nitrate (16.7 g. per 1.) required 40.2 ml. of the reduced solution corre-sponding with a total weight of thallium dibromide of 1-760 g. A method for determining the constituents of mixtures of cyanides iodides and chlorides is described. Since silver bromide is intermediate between silver chloride and iodide in solubility and adsorptive capacity for dyestuft's the method is not applicable to the determination of bromides in presence of the other halides. The titration of silver in acid solution with potassium bromide with the use of adsorption indicators yields results which compare satisfactorily with those obtained by Volhard's method. 3. A method for using adsorption indicators in the titration of halides of limited or reversible ionisation such as the thallous-thallic halides is described. Rose Bengal gives very satisfactory results in the titration of iodides in neutral or very weakly acid solution. Tartrazine and phenosafranine are well suited for the titration of silver in nitric acid solution up to an acid concentration of about normal. Phenosafranine is somewhat more restricted in its applicability than tartrazine but is preferable for work a t extreme (N/100) dilution. SUMMARY.-~. 2. REFERENCES 1. 2. 3. - Id. 1932 57 511. 4. 5. 6. K. Fajans Die Chenzische Analyss Bd. 33 pp. 161-207. A. J. Berry and P. J. Durrant ANALYST 1930 55 613. K. Fajans and H. Wolff 2. anorg. Chem. 1924 137 233. A. S. Cushman Amer. Chem. J . 1901 26 508. A. J. Berry Proc. Camb. Phil. SOC. 22 367. CHEMICAL LABORATORY UNIVERSITY OF CAMBRIDG

ISSN:0003-2654    

DOI:10.1039/AN9366100315

出版商:RSC    

年代:1936

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,

ISSN:0003-2654    

DOI:10.1039/AN9366100320

出版商:RSC    

年代:1936

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/AN9366100323

出版商:RSC    

年代:1936

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/AN9366100328

出版商:RSC    

年代:1936

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/AN9366100334

出版商:RSC    

年代:1936

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/AN9366100338

出版商:RSC    

年代:1936

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,

ISSN:0003-2654    

DOI:10.1039/AN9366100339

出版商:RSC    

年代:1936

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,

ISSN:0003-2654    

DOI:10.1039/AN9366100340

出版商:RSC    

年代:1936

数据来源: RSC