or over, while succinamic acid probably loses its water at a much lower point, and B-cyanpropionic acid rearranges to the imide at still lower temperatures. It was hoped, consequently, that the temperature at which the reaction took place, and the nature of the by-products separated, might indicate whether the course of the hydration proceeded through the cyanpropionic acid or the amide. The problem is somewhat complicated by the fact that, while only two molecules of water might be added, succinamic acid, although theoretically requiring three molecules for its quantitative production, might be continuously formed in small amount and immediately break up again into imide and water, the net amount of water, therefore, used up in the reaction being but two molecules; the production of the imide might thus occur at low temperatures and still the course of the hydration be through the amide. Then, again, the formation of imide from amic acid is partly reversible :

and succinamic acid and most of its salts change quite easily, in presence of water, to succinic.

The following reaction was taken as the basis of our laboratory work :

As the ratio of sulphuric acid to water in such a reaction is entirely independent of the particular nitrile to be hydrated, a standard acid was prepared containing sulphuric acid and water in exactly this proportion, and then the required amount measured out volumetrically. In most cases, 5 grams of the nitrile were used at a time, and the reaction carried out in sealed tubes. at temperatures varying from 100°-200°.

As expected, the imide was the chief product in all cases. Cyanpropionic acid, cyanpropionamide, and succinic acid were not found in any of the tubes. The absence of B-cyanpropionic acid and its amide is not strange, as the hydration would certainly carry the cyanpropionamide further, and cyanpropionic acid would rearrange to succinimide at the temperature of the reaction. That no succinic acid was found, however, seems to argue that the major part of the imide did not come from continuous formation and breaking down of the amic acid, for the

amic acid is so easily hydrated to succinic that in the early stages of the reaction, when small amounts of the amic acid had formed and there was still an excess of uncombined dilute sulphuric acid present, it seems difficult to understand why no trace of succinic acid resulted. Any succinic acid produced in the reaction should have been recovered unchanged, as the temperature was in no case high enough or maintained [for a sufficient length of time to cause the acid to combine with the unchanged nitrile and thereby yield the imide.

If the imide did not owe its origin to amic acid, it could not have come from the amide at all, as the temperature was too low to drive ammonia out of the amide. The only remaining explanation is that the chief source of the imide was the rearrangement of the cyanpropionic acid, and that the hydration, therefore, proceeded mainly along this line. Small amounts of succinamide and of succinamic acid were, however, isolated from many of the tubes, the presence of the latter being quite possibly due to a partial rehydration of the imide, to establish the equilibrium between the system-imide, amic acid, water. The production of imide in tubes 5 and 6, where no mineral acid was present, also seems more likely to be due to a decomposition of ammonium cyanpropionate than of ammonium succinamate. That the hy

dration should tend to complete itself upon one cyanogen group before attacking the next is quite analogous to the action of oxidizing agents, halogens, etc., upon similar straight chain. compounds.

As the question of temperature determines the relative yield. of imide, it may be said, briefly, that, for the production of succinimide, it is best to heat for about two days at 165°-175°. Below this point the hydration is slow and incomplete; above it, partial carbonization follows. For glutarimide, a temperature of 180°-200° is desirable.

EXPERIMENTAL PART.

For Succinimide.-Sealed tubes were prepared containing 5 grams of ethylene cyanide, and a mixture of two molecules of water to half a molecule of absolute sulphuric acid (= 3.6 cc. of a sulphuric acid of 1.478 specific gravity). These tubes were heated at various temperatures and for varying lengths of time, the product being a brown crystalline solid usually more or less moist according to the temperature of the reaction and the dura

tion of the heating. The method of working up the contents of the tubes was based upon the following considerations:

The bodies possibly present in the tubes after heating are the various hydration products already indicated, succinimide, ammonium sulphate, excess of sulphuric acid and water. By rubbing up such a mixture with a thin paste of barium carbonate, the excess of sulphuric acid and most of the ammonium sulphate form barium sulphate; cyanpropionic acid, succinamic and succinic acids, yield barium salts; ethylene cyanide, cyanpropionamide, succinamide and succinimide remain unchanged. If the mass be then evaporated, carefully dried, powdered, and extracted with absolute alcohol, ethylene cyanide, cyanpropionamide, and succinimide will be dissolved out, while succinamide remains for the most part insoluble. Ethylene cyanide, cyanpropionamide, and succinimide can then be separated by selection of suitable solvents, etc. The residue insoluble in absolute alcohol may contain succinamide, the barium salts of cyanpropionic, succinamic and succinic acids, barium carbonate and sulphate, ammonium carbonate, and, possibly, a little unchanged ammonium sulphate. By extraction with water, all except barium carbonate and sulphate are dissolved, and ignition of the residue will show no organic matter. Addition of sulphuric acid to the aqueous solution will then liberate the organic acids from their barium salts, when they can be removed by repeated extraction with ether. Finally, if barium carbonate be added to the acid solution after the ether extraction, and the filtrate from precipitated barium sulphate evaporated and the residue ignited, the carbonizing of the residue will show the presence of amide or of acids insoluble in ether, which may then be extracted by proper solvents until the ignition of the residue shows no organic matter.

As examples, the following tubes may be mentioned: Tubes 1, 2, 3, and 4 all contained 5 grams of ethylene cyanide, two molecules of water, and half a molecule of sulphuric acid. They were heated as follows:

No. 1, six hours at 131° to 138°.

No. 2, six hours at 142° to 155°, and then three hours at 158° to 165°.

No. 3, six hours at 154° to 162°.

No. 4, five hours at 160° to 170°, and then six and a half hours at 159° to 165°.

The contents of the tubes were worked up as already outlined. Succinimide was found in all of the tubes, the amount increasing with the rise in temperature and the duration of the heating. No. 1 contained much unchanged cyanide. Small amounts of amic acid occurred in all of these tubes, and traces of what appeared to be amide were separated in several.

No. 5 contained 5 grams of ethylene cyanide, two molecules of water, and no sulphuric or other mineral acid. It was heated for five hours at 153° to 173°. The mixture in the tube then appeared as a brownish liquid containing oily globules, and with a strong odor of ammonia. Most of the cyanide appeared un

changed, as scarcely any imide could be detected.

No. 6 carried 5 grams of the cyanide, one molecule of water, and no mineral acid. It was heated for six hours at 153° to 173°. The tube contents were then brown and semisolid. Succinimide in large amount was separated, a small amount of amic acid, traces of succinamide, but no cyanpropionamide was discovered.

For Glutarimide.-Trimethylene cyanide was heated in sealed tubes with two molecules of water and half a molecule of sulphuric acid, for five to ten hours at 155° to 200°, and the products of the reaction separated in essentially the same manner as for the succinimide tubes. Glutarimide was isolated in large amount (60 per cent. of theory), but no amide or amic acid could be detected, except the imide, the tubes containing only unhydrated cyanide. A temperature of 180° to 200° is necessary for the success of the hydration.

ORGANIC LABORATORY, HAVEMEYER HALL,

COLUMBIA UNIVERSITY,

July 1, 1901.

ON THE DETERMINATION OF CITRATE-INSOLUBLE

T

PHOSPHORIC ACID.

BY C. D. HARRIS.
Received October 12, 1901.

HE method used in the laboratory of the North Carolina Department of Agriculture up to recently was as follows: Two grams of the sample to be analyzed were washed free of water-soluble phosphoric acid. Then 100 cc. of strictly neutral ammonium citrate solution (sp. gr. 1.09) in an 8-ounce Erlenmeyer flask, was heated to 65°, in a water-bath, keeping the

flask loosely stoppered to prevent evaporation. When the citrate solution in the flask had reached 65°, the filter containing the washed residue from the water-soluble phosphoric acid determination was dropped in and the flask stoppered tightly and shaken until the filter-paper was reduced to a pulp. The flask was then placed back, loosely stoppered, in the water-bath, the temperature of which was so maintained as to give exactly 65° in the citrate flask. The flask was shaken every five minutes. At exactly thirty minutes from the time the filter and the residue were introduced, the flask was removed from the bath and the contents filtered as rapidly as possible. The residue on the filter was then washed thoroughly with distilled water at 65°. The filter and its contents were then returned to the original digestion flask, and 40 cc. nitric acid and 20 cc. hydrochloric acid were added and boiled down to about 15 or 20 cc. concentration.

The solution was diluted to 200 cc., and 40 cc., corresponding to 0.4 gram of fertilizer, were measured into a 500 cc. Erlenmeyer flask. Add 10 or 12 grams ammonium nitrate and a little distilled water. The excess of acid is neutralized with ammonia. When the contents had cooled 30 cc. of recently filtered molybdic solution were added and the flask, after securely stoppering with an ordinary rubber stopper, was placed in a Wagner shaking machine which was revolved by a hot-air motor and shaken for thirty minutes. The shaking machine was maintained at from 45 to 55 revolutions per minute, as this velocity has been found to give the maximum agitating efficiency. The flask was removed from the shaking machine and contents filtered, washed, returned to the shaking flask, and titrated.

The method of filtering the contents of the flask, after heating with citrate solution, through a funnel in which was a very thick filter-paper and platinum cone and using pressure for rapid work, was found to be unsatisfactory in three ways:

First.

The filter-paper was likely to burst and thereby necessitate another filtration.

Second. It was very difficult to fit the filter-paper in the funnel so that no air would get in around the sides and retard the filtration.

Third. It was entirely too slow for rapid work.

Having encountered these difficulties, the Hirsch funnel was