Analysis pearly scales, feeling fatty to the touch. This salt was of potash employed to determine the composition of the acid.

salt.

0.650 grammes of this salt, dried at 130° and burned with oxide of copper, gaves

0.755 grammes of CO, (inclusive of that combined with the potash, which was supposed to remain behind as carbonate), and 0.262 grammes of HO.

0.345 grammes of the salt, dried at 130°, gave by treatment with SO,HO, evaporation to dryness and ignition: 0.268 grammes of sulphate of potash=0.145 grammes of potash.

These numbers lead to the following formula and composition in 100 parts:

Deter

with

salt.

100.000

These numbers were further controlled by the determination mination of the atomic weight from the analysis of the of atomic silver salt. This salt was prepared by precipitating a weight solution of the potash salt with nitrate of silver: a white baryta apparently crystalline precipitate was formed, which slightly blackened, and then partially decomposed on boiling the mixture for a few minutes. On cooling, the salt was deposited in warty grains, which, on examination with a lens, appeared to be composed of a peculiar arrangement of needles. By careful recrystallization from a dilute solution left to spontaneous evaporation for a few days, the salt was obtained in colourless needles.

0.415 of this salt, dried over sulphuric acid, gave 0.247 of metallic silver=0.265 of oxide of silver, or in 100 parts,

6 In burning substances of this kind, and indeed all volatile bodies, or such as at a high temperature yield products of distillation which are

The formula resulting from this analysis is that of Propropionic or metacetonic acid C.H ̧OHO.

pionic

acid.

Nöll

Some years ago Nöllner obtained an acid from the fermented mother liquor of tartrate of lime, which he termed ner's pseudo-acetic acid. Nicklès analyzed the body and pseudodeduced the formula C,H,O,C,H,O,, HO, which is the acetic formula of the acid obtained by Gottlieb by the action of acid, potash or dilute sulphuric acid upon hydrocyanic ether or cyanide of ethyl (propionitrile), and which can also be obtained in several other chemical reactions. Nicklès, however, believed it to differ from the metacetonic acid. Dumas, Malaguti, and Leblanc, considered the acid of Nöllner to be identical with metacetonic acid; both acids, besides possessing the same composition, are also identical in smell, boiling points, and in the appearance and pro- not idenperties of a great number of their salts, especially the baryta salt. The subject has been again examined re- tonic or cently by Limpricht and von Uslar, who have come to propionic the conclusion that the acid of Nöllner is really different

liable to be carried forward into the chloride of calcium tube without
being fully decomposed, I have found it advantageous to use a slight
modification of the usual process. The tube is of considerable length.
Some oxide of copper is moistened with nitric acid, and heated very
strongly in a crucible, so as to get it into a hard, sintered mass, which,
when powdered, has the advantage of not being so hygroscopic as the
oxide of copper usually employed. This powdered mass is again heated,
and when sufficiently cooled, the substance to be analyzed is mixed up
coarsely with a little of it, and introduced into the tube, previously
well dried. The whole is then covered with soft oxide, heated to 200°
or 250°, and filled in out of the crucible itself in which it was heated,
as recommended by Gerhardt, and the tube being corked, it is tapped,
so as to shake the whole together. About four or five inches of the
tube in front of the mixture is then filled up with small lumps of the
sintered oxide previously heated to 200°, and put into the tube hot. The
remainder of the tube, for about six inches, is then filled with copper
turnings, whose surface has been oxydized by being exposed at a red
heat to a current of dry oxygen gas. The combustion is proceeded
with in the usual way, except that the part of the tube filled with
copper turnings, and part of that filled with the lumps of sintered
oxide, is first heated red-hot before the combustion commences, and is
kept at that temperature to the end of the operation. In this way,
any carbo-hydrogens which may be carried forward by the current of
carbonic acid and vapour of water, must pass through a kind of sponge
of oxide of copper, and over a very extended surface of oxydized
copper at a red heat, and are completely burned. When the combus-
tion is finished, the last traces of carbon may be burned away by
passing a current of oxygen gas through the tube in the way described
by Gerhardt, and which was first suggested by Laurent (Gerhardt's
Traité de Chimie, t. i., p. 35). Generally speaking, this need not be
employed, except in the case of very difficultly combustible bodies.
7 Compt. rend., xxv. 781.

8 Annal. der Chem. u. Pharm., Bd. xciv., S. 321 (1855).

tical with

metace

acid.

Diffe

rence

which

hibit in

their re

from metacetonic acid obtained from cyanide of ethyl, etc.; one of the points of difference being that the propionate of potash is separated from its alcoholic solution in a crystalline state on the addition of ether, while the butyro-acetate is uncrystallizable. The acid obtained from the peat liquor corresponds, therefore, with true propionic acid, in the circumstance of its potash salt being precipitated as pearly scales from its alcoholic solution by ether.

By repeated distillations the butyroacetic acid may be resolved into butyric and acetic acids; but propionic acid, on the other hand, may be repeatedly distilled withthey ex-out its boiling point changing. If propionate of potash or soda be distilled with alcohol and sulphuric acid, and actions. water be added to the distillate propionate of ethyle, a liquid lighter than water, and having a sort of fruit, or rather fermented fruit odour, separates. This liquid may be distilled without decomposition, its boiling point remaining constant at 101°. Butyroacetic acid, on the other hand, yields, under the same circumstances, a mixture of acetic and butyric ethers, which may be separated by fractional distillation, the former boiling at 74°, and the latter at 119°. A portion of the propionate of potash obtained from peat liquor, distilled with sulphuric acid and alcohol, yielded an ether boiling at 101°, and capable of distilling unchanged, and corresponding in every respect with propionic ether."

Treat

tween 140°

C.

The portion of the oil which came over at temperatures ment of between 140° and 160° was about three-fourths neutraloil boil- ized with carbonate of soda and distilled; the distillate ing bewas rejected. The residue in the retort was then distilled with sufficient hydrated phosphoric acid to decomand 160° pose it; the distillate was agitated with chloride of calcium, and then about one-third neutralized with carbonate of soda, and again distilled. The distillate consisted of an oily liquid, having a faint yellow tint, but becoming darker on keeping. This is also the case with nearly every peat product, no matter what pains may be taken to purify it, probably from the presence of minute traces of some body which oxidizes rapidly and blackens. The oil was neutralized by baryta water, and evaporated to dryness at a very gentle heat, a few bubbles of carbonic

See a short paper, "Observations on some of the products of the putrefaction of animal and vegetable substances", at page 202 of this number.

acid being passed through the solution in the first instance, to convert excess of baryta into carbonate. The dried residue was dissolved in water, carefully filtered, and the solution set aside for some weeks during summer in a cellar. The greater part of the salt crystallized out in the form of lustrous flattened prisms, which were unaltered by exposure to the air, and melted at a temperature a little below 100° into a clear glass. Heated to 130°, they lost 18.64 per cent., which corresponds very nearly to 4 equivalents of water.

Analysis I. 0.645 grammes of this salt, dried at 130°, and of baryta burned with chromate of lead, gave 0.730 grammes of salts. carbonic acid, and 0.264 grammes of water.

0.330 grammes of the same salt, dried at the same temperature, gave 0.247 grammes of sulphate of baryta=0.162 grammes of baryta.

II. 0.450 grammes, burned with oxide of copper, the salt having been first mixed up with phosphate of copper, gave 0.512 grammes of carbonic acid and 0.183 grammes of water.19

These numbers correspond to the following per-centage composition:

Determination

These results were controlled by a determination of the atomic weight from the analysis of the silver salt. A dilute of atomic solution of the baryta salt was treated with nitrate of weight silver, and exposed to spontaneous evaporation. After

10 It is well known that in the combustion of the baryta salts of several of the organic acids by means of oxide of copper, the baryta remains behind in the tube as carbonate, while in the case of the salts of several other acids, the whole of it does not remain as carbonate. This is the case with the acids of the series now under discussion. Lerch, in his investigation on the volatile acids of butter (Annal. der Chem. u. Pharm., Bd. xlix., p. 216), accordingly proposed to mix the salt to be analyzed, before mixing it with oxide of copper, with from three to four times its volume of ignited phosphate of copper. He found that the whole of the carbonic acid was driven off from the baryta, when baryta salts were burned under such circumstances.

salt.

some time pearly lustrous scales were formed, which were almost insoluble in cold water. They were well washed, and dissolved in boiling water. On cooling, the salt crystallized out.

0.365 grammes of the salt, dried over sulphuric acid and ignited, left 0.201 grammes of metallic silver=0.216 grammes of oxide of silver. These results lead to the following numbers:

Butyric acid.

Constitution

rate of

baryta.

These numbers represent the formula C,H,O,,HO,

which is that of butyric acid.

The salt of baryta with four equivalents of water, which I obtained, corresponded with one of the salts described of buty- by Chancel, and which appears to have been also obtained by Pelouze and Gelis. Chancel stated that butyrate of baryta crystallizes either with two or four equivalents of water of crystallization, according as it is obtained from hot or cold solutions. When it crystallizes in the cold, it forms long flattened prisms perfectly transparent and containing 18.8 per cent. of water, or four equivalents. These crystals melt to a transparent liquid without losing weight when heated to a temperature below 100°. Crystallized from warm solutions, the salt contains 10.5 per cent. of water, or 2 equivalents, and does not melt at 100°. I had not enough of material to succeed in the preparation of the salt with two equivalents of water, having only obtained anhydrous crusts in every attempt which I made.

According to Lerch, butyrate of baryta appears in two different crystalline conditions, and in each is anhydrous; the one consists, as Chevreul has described, of mother-ofpearl-like plates and flattened flexible prisms; the other of hard granular crusts. The latter may, however, be converted into the former by repeated recrystallization. In either condition it did not melt at 100°. Lerch further states that he could not obtain butyrate of baryta other than in an anhydrous form.

It is difficult to account for the different results obtained by Chancel and myself, and by Lerch. The same difficulty occurs in the case of propionate of baryta, if the salt of Keller, with nine equivalents of water, be