THE ANALYST. THE EXAMINATION OF TURPENTINE AND TURPENTINE SUBSTITUTES. BY J. H. COSTE, F.I.C. (12end at the Meeting, May 6, 1908.) Preliminary. IT is generally known that the various species of coniferous trees yield, when bled,” more or less considerable quantities of oleoresin, from which by suitable means a volatile spirit or oil of turpentine can be distilled. Of all these varieties, American turpentine, principally obtained from the Georgia and loblolly pines-Piitus ilzLstraZis and Pims tmh-is the best known, and commands the highest price, its principal use being a diluent or solvent material for paints and varnishes. This material is usually distinctly destro-rotatory, and has a strong and not disagreeable odour. I t is completely volatile at ordinary temperatures, or at most leaves mere traces of osidised products.Its principal constituent is dextro-pinene. American turpentine also contains dipentine and various oxidation products. Furfural appears to be usually present in traces. Much of the American turpentine sold to-day does not conform to the foregoing description. I t is of an unpleasant acrid colour, at times lcevo-rotatory, and, judging from its behaviour on distillation, contains but little of either dextro- or lzvo-pinene. This alteration in character which has occurred of late years is shown in Table II., and is due to one or both of the following causes : (1) The forests of Georgia and Carolina, having been worked in a very reckless manner, are now exhausted, and, consequently, new districts towards Florida have been opened up, the turpentine from which is of a diflerent character.(2) The methods of distillation have been varied, and not only is the volatile (‘ spirit ” distilled from the oleoresin in the old-fashioned manner, but the wood and stumps of the pines themselves are submitted to a process of distillation, which yields a crude turpentine, known locally as “ stump spirit ” or Whichever cause is at work-and it appears most probable that both are contributory--there is no doubt that much of the turpentine shipped to Europe from the United States is of a very different character from that which a few years ago was recognised as typical American turpentine. I t is certain that wood turpentine as such can only be obtained with difficulty in London, although much of it is distilled in America.Some is doubtless used in the States for varnish-making und painting, but it appears very probable that a great deal is shipped to Europe wood turpentine.”220 THE ANALYST. either as a substitute for (' box " turpentine or mixed with it as an adulterant. That this is a very general opiniou amongst analyst8 may be gathered' from the various methods suggested for its detection. Adulteration of American turpentine with the more volatile fractions of petroleum appears to be less frequent than was the case before "wood turpentine" was available for the purpose. Occasionally, however, cases of such adulteration come under one's notice, more particularly, perhaps, in the case of varnishes than in turpentine supplied as such under contract.Russian and Swedish turpentine from Pinus SyZvestyis and P. ledebo?wdii are now refined, with the object of removing the unpleasant persistent odour which is so characteristic, and sold for painters' use under various names. They have a some- what higher specific gravity, boiling-point, and refractive index than American turpentine, and are strongly dextro-rotatory. Their principal constituent is sylwestrene. One drop of sylvestrene dissolved in a little acetic anhydride gives an intense blue colour with one drop of concentrated sulphuric acid (Russian turpentine itself appears not to give this colour). German turpentine is of very similar character, as also is much of that from Finland. French turpentine, from P. pinastey, is similar to American, but lmo-rotatory, owing to Imo-pinene being its principal constituent.According to Long (Jozwn. Amer. Chenz. Soc., 1894,16, 844-847), spruce turpentine, from P. gtabra, is strongly lawo-rotatory-from 63.1" to 70.3" in a, 200-mm. tube. In other respects it appears to be like ordinary American turpentine. Various turpentine substitutes, sold under names more or less similar to turpentine, are mostly petroleum products mixed with small quantities of Russian turpentine, or waste terpenes from other industries, in order to give an odour of turpentine. The boiling-point has a very wide range, and the refractive index is low. These materials mill be considered in detail later. The specific gravity is usually between 0.78 and 0.82. Methods of Examination. Many processes have been suggested for the examination of turpentine.The following are mentioned, not as being novel, but as having been found useful, during a long experience, in aiding one to form an opinion as to the nature and quality of samples of this material. Speci$c Gyavity (D ~ """ C.).-This varies between from 0.8G and 0.871, A 15-5" lower value may be due to petroleum or shale oil, a higher to resin spirit, Russian turpentine, or old oxidised turpentine. Refyactive Index (n,20) varies from, say, 1.4690 to 1.4720, and is influenced similarly to specific gravity. This can be very conveniently determined by means of either the hbb6 or the Zeiss refractoineter. Temperature correction -f 1" C. T .00037. Optical rotation affords no very direct information as to quality, but is frequently useful in forming an opinion as to the nature of turpentine wheu considered in con- nection with other figures. I t is now geneyaZZy much lower than mas the case ten years ago.Temperature correction +_ lo C. T -00066.THE ANALYST. 221 Initial boiling-point 157" and 160" C. ... ? ) ,, 160" and 165" C. ... 1 Y ,, 165" and 170" C. ... ' 9 ,) 170" and 175" C. ... #, ,, 175" and 180" C. ... FrnctionaZ DistiZZution.-The following table will give some idea of the relative amount of separation obtained by distillation of 100 C.C. from (1) a three-bulb Laden- burg flask, (2) EL round-bottomed flask with a plain, Wurtz tube about 6 inches long, and (3) a Young's rod and disc " column : 44 C.C. 14.5 C.C. 45 C.C. SO ,, 54.5 ), 31 7 ) 9.5 ,, 14 > Y 8.5 Y ? 4 ,, 5 9 , 3.5 Y 7 3 ,) 3 9 ) 3.0 Y9 TABLE I.-100 C.C.AMERICAN TURPENTINE (ROTATION IN A 1 -DECIMETRE TUBE +16" 35' n1)?O0 1.4686). A rn oun t of' Dis- t ill:t te. L'einpcra. ture "C. 159 159-3 159.6 160 160.8 161.5 164.0 168.0 180 1. Rotation 1-Ueci- metre Tube. 111 iL 20" 8' 20" 0 19" 45' 19" 34' 18" 57' 18" 12' 17" 22' 15" 28' 11" 29' 12D2"O. 1.4662 1.4663 1.4666 1-4666 1 *4668 1.4674 1.4680 1.4683 1.4704 l'empera ture "C. 159.6 160.3 160.7 161.3 162.1 164.7 165.2 168.9 180.6 2. Rotatioii i n a 1- Decimetre Tube. 19" 40' 19" 41' 19" 35' 19" 7' 18'35' 18'7' 17" 1' 15" 34' 11" 43' Distillate between 1.4660 1.4663 1.4667 1.4668 1.4671 1.4673 1-4677 1.4684 1 -4 704 Tempera ture "C. 158.1 158.6 159.0 159.5 160-1 161.2 162.6 166.4 180.6 3. Rotation in a 1-Deci- metre Tube.20" 58' 20" 39' 20" 21' 19" 51' 19" 40' 19" 2' 17" 57' 15" 54' 10' 30' 1.4656 1.4660 1.4661 1.4664 3.4668 1.4671 1.4675 1,4679 1.4702 ..* 1 90.5 ,, Total ... ... ... The Ladenburg flask is very convenient, and the separation is not very different from that effected by the "rod and disc" column. A three-bulb Ladenburg flask of 180 C.C. capacity to the bottom of the neck is adopted by the author as a standard flask for this and similar distillation tests, using 100 C.C. of liquid. The time of dis- tillation is about twenty minutes. A No. 4 Anschutz thermometer is of convenient size and range. I t should be standardised with brombenzene (boiling-point 155" C.) and aniline (boilingpoint 184.1" C.), and should be carefully matched to see that any mercury condensed on the upper part of the tubeis united to the main volume before using.222 THE ANALYST.The normal constituents of American turpentine are stated to be : tl-pinene . . . ... ... Cymene* ... ... ... Dipentine (i-limonene) . . . Resinous matters . . . ... E. P. 155-6" 178.0" 175-6' - n ,,20° 1.4652 1.4731 1.4926 (at 13.7") Very high 15.5" 15.5 D -u c. 0.861 0.848 0.8722 (at 0') Very high Rotation in a l-Dcm. Tube. 40" 0" 0" - Of these pinene is the principal and characteristic constituent, dipentine being the result of overheating in the process of distillation, and cymene and the resinous matter the products of oxidation due to keeping. A pure, carefully prepared, and fresh turpentine, therefore, should clearly yield on distillation no constituent having a lower boiling-point than 155" C.; the greater portions should distil at tempera- tures very close to 155' C., and all, with the exception of a very small residue, should distil below 180' C.None of the distillate should have a lower refractive index than that of pinene, and the first portions should have the lowest refractive index and the highest optical rotation, as the optically active constituent (pinene) is also the constituent of lowest boiling-point. This view is confirmed hy the detailed results in Table I. I t is usually desirable to note the initial boiIing-point, and if below 155" C., to collect separately the fraction which is distilled up to that temperature. If it contains any water, this should be tested for acetic and formic acids, the presence of which would be evidence of old turpentine.The oily portion is, if necessary, dried, and its refractive index and optical rotation determined. If the refractive index is below that of pinene, and the optical rotation is lower than that of the original sample or of the next fraction, adulteration with an optically inactive liquid of low refractive power-e.g., light petroleum-is indicated. Confirmation of this may be obtained by Armstrong's method (vide infra). I n any case, the first 5 or 10 per cent. of distillate from 155" C. upwards are collected apart, and the refractive index and optical rotation determined. These should conform to the above statie- ment. The amounts of distillate obtained at each 5" from 155" to 180" C. are noted, and it is useful also to observe the temperature at which each 10 C.C.from 10 to 90 C.C. comes over. A good American box turpentine should, in the opinion of the author, comply with the following specification : 100 c.c., on distillation in a three-bulb Ladenburg flask of 180 C.C. capacity to neck, shall yield no distillate before the temperature reaches 155' C., and not less than 70 C.C. between that temperature and 160" C., and a total distillate of not less than 95 C.C. up to a, temperature of 180" C. The amount of * A sample prepared by t h e anthor from camphor by t h e action of 1'205 had ?in'LO"=l -4896, the correction for +_1" bring T.0004, and boiled from 174.5" t o 177" C. D 1 ~ ' ~ " o . 8 6 ~ ~ . 1d.d"THE ANALYST. 263 distillate up to 180" C. is really higher than that sbserved by about 1 per cent.(varying with the length, etc., of the condenser), when allowance is made for wetting the condenser and for the flaskful of vapour left behind. Free dcitZ.-It is advisable to titrate 10 C.C. diluted with alcohol, and to note the volume of :c alkali required to neutralise the free acids-acetic, formic, and probably abietic-using phenolphthalein as indicator. Steam distillation occasionally gives useful results, but generally fractional distillation will impart as much information with less trouble. For the direct determination and separation of petroleum spirit in turpentine che author is of opinion that the process of polymerisation and sulphonation pro- posed by Armstrong (Joz~rn. SOC. Chmn. Ind., 1882, 1, 480) is in every way preferable 80 the various drastic and even pyrotechnic processes suggested by later workers. Briefly, the method is to polymerise, carefully avoiding rise of temperature, with diluted sulphuric acid (2 volumes acid; 1 volume water).Any petroleum spirit or unaltered oil of turpentine is separated from the less volatile polyterpenes by steam dinhillation. The distillate is then treated with stronger acid (4 volumes acid; 1 volume water), and the portion volatile in steam collected. If this does not exceed 5 per cent. of the original volume of turpentine, it is probably cymene which is normally present in old turpentine, and is formed to a small extent by the action of sulphuric acid. This can, if desired, be identified by its properties : - D 13'70 C. 0.8619, B.P.175-6" C., nJ3.7 1,4926, optically inactive. 4" I t is also completely sulphonated by treatment at 50" to 60" C. with diluted Nord- hausen acid. Any excess over 5 per cent. of unpolymerised residue would he petroleum, which, after removal of cymene by sulphonation, can be detected by its low specific gravity, wide range of boiling-point, low refractive index, and its behaviour to sulphuric and nitric acids. The amount obtained in this process represents the minimum of adulteration with mineral oil; shale oil is too much acted on by sulphuric acid to be detected in this manner. As small a quantity as 100 C.C. of turpentine will with care give reliable results by this method. A qualitative test for wood tziipentiite, recommended by Conradson ( J o ~ L ~ . SOC.CI~enz. I n d . , 1897, 16, 519), Flath (ANALYST, 1908, 27), and others, is agitation with an equal volume of a saturated solution of sulphurous acid. Wood turpentine assumes a yellow colour, and Russian and Swedish turpentines a yellow-greenish cobur. This test certainly more often gives a yellow coloration with the newer type of turpentine than with the older more volatile kind. Turpentine should not leave a stain on paper which has been coated with it. The paper should take ink freely after hanging up to dry. The comparative volatility of samples can be determined by exposure of 0.5 or 1.0 c.e. of each in similar flat dishes to the air of the laboratory, noting: from time to time the loss in weight which occurs. If time and loss are plotted a curve is obtained, which does not begin t o flatten until nearly all the liquid is volatilised.The various substitutes show 8 falling off in the rate of volatilisation much sooner than genuine turpentine. American or French turpentine remains white. All wood oils are said to give this yellow or green colour.224 THE ANALYST. The detailed results of the examination of a few selected samples examined during the period beginning in 1899 are given in the following table : TARLE II.-RESULTS OF THE EXAMINATION OF SAMPLES OF TURPENTINE SUPPLIED AS GENUINE AMERICAN TURPENTINE. DiHtillation of 100 C.C. Properties of First 10 C.C. of Distillate. u; 0 co H 0 4 0 t- u' 3 s 0 0 +l d 0 L- H 0 +a 0 (0 m 1. 2. 3. 4. 5. 6 . 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 1s. 19. 20. 21. 22. 23. 24.25. 26. 27. 28. 29. 30. - 0.8dSe 0.8611 0.8677 0.8688 0'8670 0.8650 0.8685 0.8697 0'8504 0.8693 0-8680 0.8715 0.8706 0'9141 0-8670 0.8682 0.8712 0'8668 0.8662 0.8676 0.8670 0.8659 0.8707 0'8701 0.8677 D.8700 3 8655 1.8658 1.8664 1.8670 80 -0 47'0 i8.O 72.0 85.0 93.0 92.0 76'0 42.0 83.0 20.5 63-0 30.0 25.0 81.0 51 *o 13.0 57.0 18.0 92'0 i 5 . 0 10.0 30.0 i 9 . 0 70.0 18.0 j2.0 17'0 38-0 ZO.0 - 13.0 26'0 16.0 16.0 7.0 4.0 3.5 18.0 25.0 10'0 32 *o 125.0 35.0 30.0 11.0 39 -0 18'0 19.0 33.0 66'0 4'0 15.0 72.0 58.0 13.5 23.0 42 .O 30.5 68.0 6.5 2.0 9.0 1.0 3.0 1.0 1 * 5 5% 9 *o 2.0 35 '0 5 . 0 9.0 9.0 3.0 4.5 3.0 2.0 4-0 5 '5 1 *5 4'0 6.0 4.0 2.5 3.0 4.5 2.5 7.0 1 .o - 1 -0 4.0 1'0 2.0 2'0 - - - 6 -0 - 7'*0 2.0 8.0 4.0 1.0 1.0 2 -0 2.0 1 -0 3.5 1.5 2'0 2.5 1 '0 0 '5 1 *5 1 *5 2.0 1 -0 - 97.0 89.0 97.0 94-0 96.0 97.0 97.0 99-6 85.0 95.0 96.5 96.0 85.0 71.0 97.0 97.0 94 '5 97.0 95-0 94.0 97-5 96.0 33.0 36.0 36.5 37.5 36.0 )6..5 36.0 16.5 - - +2" 6' $24" 50' - - - - - - - - - - -1-6" 50' i-3" 57' +7" 59' -0" 44' +11" 19' +5" 8' +lo 39' + 4 9' -29" 44' +2" 4' t-2" 10' f 30 0' t 6 " 15' t 2" 45' t 4" 0' -0" 10' - - - __ - - - - - - - - - Helow 155.0" 154.0" 155 -0' 155.0" 1 5;-0" 165.5" 154.5" 159.0" 158.0" 157.0" 158.5" 158.5" 157.5" 158.8" 156 5" - - - - - - - - - -_ - - - - - - 1.4689 1.4686 1 '4686 1'4694 1.4683 1.4682 1.4680 1.4699 1.4685 1.4686 1.4681 First 1 -4680 2.4680 1.4699 1.4674 1.4676 .. - -_ - - - _._ - - - - _ _ - 4-9" 0' +5" 55' +11° 20' +I0 45' +14c 0' +7O 2' - 33" 11 ' +4" 15' i per cent. + 5 O 30' +go 5' +5O 0' +i" 0 + 2" 50.- - - - - - - -_ - - - - - - - - 1 *4709 1.4713 1 '471 4 1'4710 1.4703 1'4714 1 *4699 1.4693 1'4715 1.4707 1-4702 1.4699 1.4699 :4704 . -4703 . *4700 Some few samples call for special remark : No. 2 : The low specific gravity and small amount of total distillate point to the presence of petroleum. After polymerisation and sulphonation, 8.6 per cent. of a 15.5" liquid was left, with D ~ 0 . 8 1 6 3 , and range of B.P. 150" to 220° C. ; it was evidently petroleum. No. 9 : -4gain, the specific gravity and result of distillation point to adulteration with petroleum. The residue, after treated by Armstrong'sTHE ANALYST. 225 15*53 method, was 15 per cent. by volume, and had D 1m 0.7879. No. 13 : On further distillation, 10.5 per cent. by volume was obtained up to 225" C .The original liquid contained at least 12 per cent. of a liquid unattacked by either sulphuric or nitric acids, with D 15'5" 0.8376, and range of B.P. 160" to 210". No. 1 4 : The high 15.5" specific gravity and small amount of distillate indicated the presence of resin oil or old turpentine, which was confirmed by the high acidity figure. The residue in the 15.5" distillation flask was of D 1m 1.0048, and required 2.8 per cent. of KOH to saponify. Having regard to the fact that 7 C.C. distilled below 155" C . , it is fairly certain that this was old turpentine containing acetic and formic acids. No. 15 : This, after polymerisation (not sulphonation also), left a residue of 2.2 per cent., with n,20° 1.4822. This was cymene frorn genuine turpentine.Nos. 18 and 30 : The change of sign of the rotation is remarkable as indicating the presence of two active constituents, one dextro and the other l w o , of different boiling-points. This, like all the other samples in the above table, was sold as American turpen- iine. The free acid in the above samples was on the average equal to about 0.3 per cent. of abietic acid, corresponding to 0.8 to 1.0 C.C. & KOH for 10 C.C. of turpentine. The author is of opinion that only those samples in the above table which comply with the specification previously mentioned can be considered as satisfactory. The results of the examination of a few samples of Russian turpentine will suffice to indicate the general character of this material and its marked difference from American turpentine.I t is worthy of note that the rubefacient properties of Russian turpentine are mom marked than those of American turpentine, so much so as to render it an unpleasant material to handle for cleaning or polishing purposes." No. 23 I t is probably either French or spruce turpentine. No. 1. No. 2. ... ... ... ... ... 0.8662 ... 0.8837 D -To n 20' ... ... ... ... ... 1.4740 ... 1.4750 15.5" 15 5 Optical rotation (1-dcm. tube) ... ... + 14" 29' ... + 16" 20' Percentage distillates from Ladenburg flask between 145" and 160" C. ... ... 4 (157Oand 16OOC.) 2 160" and 165" C. ... ... 12 16 165" and 170" C . ... ... 43 36 170" and 175" C. ... ... 20 18 175' and 180" C . ... ... 11 7 180" and 185" C. ... ... 3 4 185" and 190" C. ... ... 2 2 No. 1, on polymerisation by Armstrong's method, yielded 14.3 per cent.(volume) 15.5" of unaltered matter with D 15.30. 0.8496 and ~ Z D ~ O ' 1.4778. Distillates : 146" to * This property of Riissiaii tiirpentiiie was indicated to a niarked and painfiil degree in coniparative .r?xpcrinients made, under the i i i ~ t h o ~ ' ~ supervision, by q)plyiiig Rnssiaii and Anierican turpentine :LR poultices to the wrists, with i l view of investigibtillg i i con'iplaint from workinen using Russian twpciitinr.296 THE ANALYST. 160" C., 2-5 per cent. ; 160" to 165" C., 3.7 per cent. ; 165" to 170" C., 11.2 per cent. ; 170" to 175" C., 56.2 per cent. ; 175" to 180" C., 20 per cent. ; 180" to 185"C., 2-5 per cent. ; total, 96.1 per cent. On sulphonation of this with Nordhausen acid, a volume corresponding to 4.3 of the original turpentine was obtained, having a refractive index 1.4528.Evidently the residue of the polymerisation was principally cymene, which is said to be largely present in Russian turpentine, with about 5 per cent. of petroleum. Such adulteration with petroleum has frequently been noticed by the author in Russian turpentine. When turpentine is exposed to the air in vessels sufficiently closed to prevent evaporation, oxygen is absorbed and ozone (or hydrogen peroxide)" formed. This causes the bleaching of the cork often observed in such cases. Considerable oxida- tion also occurs; aoetic and formic acids are formed, with cymene and resinous products. The physical changes occurring are indicated by the following figures : TABLE III.-AMERICAN TURPENTINE.15.5" D 15rpC. ..- Distillates : 160" C. 160" to 165" C: 165" to 170" C. 170" to 175" C. 175" to 180" C. Total ... Rotation in a l-dcm. tube Refractive index ... Viscosity (Red- wood's visco- meter) ... Free acid (c.c. per- 100 C.C. liquid) . . . October, 1903. 0.8686 92.5 3.0 - - - I. October, 1904. 0.9040 29 29 11 5 3 November, 1906. 0.9540 7.0 12.0 23 8 4.5 54.5 - - - - February, 1905. 0.8717 51 37 6 1 1 96 4" 4' - - 0.5 11. October, 1906. 0.9091 14 36 10 10 - 70 5" 1' - - - March, 1908. 0.9270 18 24 14 6 6 68 5" 0' 1.4802 46 sec. (Fresh turpen- tine, about 30 seconds) ; soluble acids, 8.4 ; insol- uble, 134. * Iiingzett has shown that hydrogen peroxide, and not ozone, is formed by the atmospheric This certainly appears to be the case, although the contrary statement has oxidation of turpentine.been made on good anthority.THE ANALYST. 227 Old turpentine leaves a, greasy stain on paper. The residue after distillation of 11. had a specific gravity of more than 1.0, and refractive index 1-5077, and resembled resin spirit,. If kept in a well-corked bottle, turpentine is not altered appreciably on keeping for a year. I t will be noticed that the refractive index and boiling-point of the first fractions obtained on distillation of turpentine, as shown in Tables I. and II., do not agree closely with those of pinene, but one or other of these figures is always higher than one would expect if the mixture distilled contained constituents boiling between 155°C. (pinene) and 175" C. (cymene).The existence of at least one other normal constituent of turpentine appears to have been firat presumed by v. Bayer (Ber., 1896, 29, 1923), who obtained from the products of oxidation of French turpentine, by means of a neutral solution of permanganate of potash, as indicated by Tiemann and Semmler (Ber., 1896,29, 529), an acid which he called B-pznonic acid, to distinguish it from the pinonic acid obtained by those authors, and another acid to which he gave the name nopic acid, isomeric with pinonic acid, but difTering from it in not containing the keto-group. Pure pinene did not yield nopic acid on oxidation. The acid was, therefore, presumed to be the oxidation product of another terpene, to which the name nopinene or p-pinene was given. Recently Wallach (Annulen, 1907, 357, 49-71) has, by suitable means, prepared from nopinone, a derivative of nopic acid, a, terpene possessing the following properties : Boiling-point 158" C.D20" 0.8630 [.ID + 15*93", n,20° 1.4699. This is apparently not the /3-pinene of French and American turpentine, but is closely allied to it. The existence of a turpentine possessing properties very similar to this P-pinene or its laavo-isomer is certainly indicated by the properties of the distillates from American turpentine. Year. 1899 I.. 1900 ... 1901 ... 1902 ... 1903 ... 1904 ... 1905 .._ 1906 ... 1907 I . . NO. Examined. 2 18 14 13 20 26 54 78 35 CONTKACT 1. TABLE IV. CONTRACT 2 (and Casual Samples). Percentage Passing Specification. (a) 100 67 93 92 100 8 11 10 37 NO. Examincd. 26 24 35 38 39 63 61 52 59 Percentage Passing Specitication. ( a ) 54 83 97 97 85 11 20 31 15 The terpenes are a, class of bodies very liable to change in structure, and it must be borne in mind that the methods of distillation, even when the oleoresin itself is distilled, are such as to yield not merely the volatile hydrocarbons of the turpentine,228 THE ANALYST. but secondary products of the nature of dipentine, cymene, etc.Probably, if distilla- tion with low-pressure saturated steam or distillation under reduced pressure were adopted, much more uniform results would be obtained. When pine-wood is distilled more or less destructively either alone or with superheated steam, more complex products must be expected. The general alteration in character of turpentine examined by the author during the last nine years is indicated in the following summary, &owing the extent to which turpentine supplied under two different contracts, and some casual samples (usually of superior quality to those supplied under contract), complied with the specification mentioned previously-Le., to yield 70 per cent.distillate between 1 5 5 O and 160" C. and 95 per cent. up to 180' C. (column a, Table IV.), and also the percentage in each year yielding 60 per cent. up to 160" C., and a total of 95 per cent. (column b). I t will be seen that in both cases the year 1904 marks the change which has been indicated by the great decrease in the amount distilling between 155' and 160' C., and a corresponding increase in the next fraction, 160' to 165" C.If the somewhat strongly expressed belief in the sulphurous acid colour-test is well founded, a great deal of wood turpentine must find its way to the United Kingdom; but in the author's opinion such tests are not very reliable unless the chemistry of the process, which in this case appears to depend on the presence of aldehydes and ketones, is well understood. Under the circumstances, it appears very difficult to state more of a sample than whether (1) it is adulterated with petroleum or hydrocarbons other than the terpenes and their allies; (2) it is fresh and in good condition, or old and badly kept ; (3) its general properties are such as to establish the presumption that it is freshly distilled American (or French) box turpentine, or a second-rate but simiIar product ; (4) it is Russian turpentine, or a similar mixture of terpenes ; or (5), what- ever its nature, whether its volatility at ordinary temperatures is such that it is probably sufficiently good for use as a diluent for paints and varnishes.This last point lertds to the question of turpentine substitutes. Turpentine Substitutes. The general character of these materials has already been indicated ; as to their utility, much diversity of opinion exists. The volatility of the best of them at ordinary temperatures is inferior to that of American turpentine, a very marked flattening of the curve being noticed. This volatility is probably, however, in the better kinds suficient for most purposes, and the materials are much cheaper than turpentine. I t is sometimes said that turpentine leaves a residue possessing some occult properties of a highly satisfactory nature.This residue cannot be very different from ordinary colophony, which is not a particularly good material in paint w varnish, and which, if it were so, eould easily be added to the substitutes. Turpentine may, by its power of forming ozone or hydrogen peroxide on evaporation, aid the oxidation of linseed oil, and in this respect be superior. Good turpentine has a more steady rate of evaporation than the best substitutes,THE ANALYST.. 2222 and it is also not of so persistent an odour. I t is dificult to obtain definite facts from trade sources. Some makers of high-class varnish point out thah with a cheaper solvent they can for the same price give better gums in a varnish, and maintain that no differenee in working exists between turpentine varnishes and those with good substitutes.Others say that for high-class work nothing is equal to American turpentine. The substitutes are certainly much used in varnishes, and appear to be quite satisfactory so far as the appearance of the coat is concerned. This might be expected, when one considers that in both paints and varnishes the solvent is only added to make the material workable, and is not a permanent constituent of the film applied to the surface to be protected or decorated. For wax-polishing of floors substitutes are found to be quite satisfactory, I t is certain that American turpen- tine will not be entirely ousted from the field until the various substitutes are made of a more uniform character.The results obtained in the examination of several samples of substitutes am given in the following table : TABLE V. d 7, 1 2 3 4 5 6 7 8 9 10 11 12 13 1 4 0 0 2 1::;: - I P n 0.8210 0.8816 0'8204 0'8088 0.7981 0.8073 0'8011 0-8030 0.7977 0'7896 0'8004 0'8144 0'8079 0 *7 9 8 4 1'4504 1.4.5 10 1'4493 1'4493 1.441s 1.4488 1.4454 1.4457 1'4426 1 -4394 1.4447 1.451% 1'4492 - ci 0 9 3 0 Y 3 7 3 3.0 * 2.0 1.0 14.0 13.5 12.5 11.5 14.0 11.0 24'0 2.0 1'5 2.0 Distillation of 100 C.O. 3.3 * L 1.2 11.5 11'5 7 . 5 9.0 S *5 12.0 1s.o 14.0 3 .O 13'0 5.7 12.0 10'0 1% 11.6 9.0 6.5 7 -5 9.5 11.5 11.6 11.0 5 . 5 5.0 -.- d m (0 3 0 + 0 0 9 4 4.0 6 . 5 - 4.7 9.0 s *5 7.0 3.5 8.0 9.0 9'0 12.0 5.0 9.0 5.0 2.0 6-0 9'5 8.0 7.0 5.5 2.5 9.0 i . 5 S.5 4.0 9.0 r r 1 . 3 9.0 T ' O 11.0 20 '0 5-0 6.0 3.5 5.0 7.5 5-0 4'5 11'5 4.5 s.5 8'0 10'5 9-0 30.0 7.0 '7.0 8.0 10'0 7.5 7.0 5'0 10'0 5.5 9.0 - m c 35.0 38.0 39.0 68.0 75.0 74.0 53.5 65.0 64'0 75.5 92.5 69 *o 30.0 57.5 Of these materials, Nos. 1, 2, 3, 4, 5 , and 11 appeared to be suitable for use in the preparation of oil paint. That marked No. 11 is a particularly satisfactory material, and is of very uniform character. I t will almost invariably pass the following specification: The specific gravity to be 0.790 to 00810 at 15.5" C. ; 100 C.C. on distillation in a, three-bulb Ladenburg flask of 180 C.C. capacity to neck shall yield not more than 2 C.C. of distillate below 140" C., nor less than 85 C.C. between 140° C. and 180" C. The somewhat low flashpoint does not appear to * Amount not noted.230 THE ANALYST. introduce any real element of danger. This material is largely used for wax floor- polishes and for high-class varnishes.::: I t will be seen from the properties of these materials that any considerable proportion could not be mixed with turpentine without being detected by the lowered specific gravity and refractive index. None of these materials left any stain on paper. CHEMICAL DEPARTMEST, LONJIOS COCSTY COUSCIL.