Of these, 1 is the total globulin extracted by brine from one portion of oil-free meal; 2, 3, 4 and 5 are fractional precipitates from another similar extraction; 6, substance precipitated by saturating the salt extract with sodium chloride; 7 and 8, substance soluble in saturated sodium chloride solution, and 9, that precipitated by cooling an extract made with a one and a half per cent. salt solution heated to 60°.

These results show that the most abundant proteid of the sunflower seed consists of a single globulin, and that the proteid precipitated by saturating with sodium chloride contains the same amount of nitrogen as the proteid soluble in a saturated solution of this salt. As Vines stated that the substance of the aleurone grains was soluble in a saturated salt solution after treatment with alcohol, while after treatment with ether it was insoluble therein, although soluble in ten per cent. salt solution, we thought that possibly by treating our meal with alcohol we might remove some substance, perhaps an acid soluble in alcohol but insoluble in ether, which might be the cause of this peculiar behavior of the proteid. We accordingly extracted a quantity of sunflower meal with alcohol of 0.820 sp. gr. and in order to determine whether acid had been removed we attempted to titrate a portion of the extract with a one per cent. solution of potash. On adding the alkali a colored precipitate resulted which rendered the indicator (phenolphthalein) useless. The attempt was then repeated, omitting the indicator. When the potash solution was added a bright chrome yellow color resulted, which gradually increased with the formation of a precipitate as the quantity of potash was increased. With a larger excess of potash the precipitate redissolved. This reaction we found to be due to helianthotannic acid (Ludwig and Kromayer, N. Br. 99, 1 and 285). The results of our investigation of this acid will be given in another paper.

Having now found a very delicate test for this acid, we applied it to our preparations of globulin and obtained a strong reaction in every case. It was therefore necessary to remove this acid from the meal before attempting to obtain the proteid and accordingly the extraction of the meal with alcohol was continued. It was, however, practically impossible to remove the acid so completely as to obtain no yellow reaction when the extract was treated with potash.

The meal which had been nearly freed from this acid was

washed with ether and air dried. 100 grams were extracted with ten per cent. sodium chloride brine and the filtered extract saturated with salt. An abundant precipitate separated, just as with meal which had not been treated with alcohol. This was filtered off, dissolved in ten per cent. brine and again precipitated by saturation with salt. This precipitate was again dissolved in salt solution, filtered perfectly clear and dialyzed. The globulin which was thus precipitated was filtered out, washed with water and alcohol and dried over sulphuric acid. This preparation, 10, weighed 7.4 grams and had the following composition :

The saturated sodium chloride solutions filtered from the two precipitations of 10, were united and dialyzed until free from chloride, the resulting precipitate was filtered out and treated as 10 had been. Preparation 12 was thus obtained, which on analysis gave the following results:

As both the preceding preparations were found to contain detectable quantities of helianthotannic acid another attempt was made to prepare some meal which should be practically free from this acid.

One hundred grams of meal were therefore extracted in a Squibbs percolator with alcohol of 0.820 sp. gr., the whole being kept at 65° C. until 1500cc of extract were obtained.

The temperature was then raised to 75° and the extraction con

tinued, about seven liters of alcohol being passed through the meal. The last two liters were evaporated and left a residue weighing only 0.28 grams.

The meal residue was air dried and extracted with ten per cent. sodium chloride solution. The extract was then filtered clear and saturated with ammonium sulphate, the precipitated proteid filtered out, dissolved in brine, the solution filtered perfectly clear and dialyzed.

The proteid was thus precipitated in large spheroids and was filtered out, washed with water and alcohol, dried over sulphuric acid and found to weigh 15.5 grams, preparation 13. This substance was freer from coloring matter than any before made, and had the following composition:

This preparation, which was very nearly white in color, dissolved after drying almost wholly in ten per cent. sodium chloride brine at 20°, giving a solution slightly tinged with greenish brown, which on dilution yielded an abundant precipitate that on warming, while suspended in the diluted solution, redissolved completely and again separated on cooling in spheroids, and on settling united to a coherent layer.

Solutions in ten per cent. sodium chloride brine behaved as follows:

When saturated with magnesium sulphate at 20° or sodium sulphate at 34°, the proteid was completely thrown out of the solution. When saturated with sodium chloride it was partly precipitated.

With mercuric chloride, picric acid or tannic acid a heavy precipitate was produced.

With minute quantities of nitric, sulphuric, hydrochloric or acetic acid the globulin was precipitated.

In pure water this preparation formed a plastic mass, but none dissolved.

In water containing a minute quantity of acid it dissolved readily and completely.

With the xanthoproteic, Millon's, biuret and Adamkiewics' tests the usual proteid reactions were obtained.

When dissolved in ten per cent. sodium chloride solution and tested for heat coagulation point in the usual manner a turbidity formed at 90°, and a flocculent coagulum began to separate at 93°, increasing as the temperature was raised toward 100°. After heating some time in a boiling water bath a considerable coagulum formed, yet a large proportion of the substance still remained in solution, as shown by the voluminous precipitate produced on adding acetic acid to the solution filtered from the coagulum.

In composition and reaction this preparation agrees with the globulin edestin except that a part is precipitated by saturating its solutions in brine with sodium chloride. In composition the part precipitated by saturating with salt and that remaining in solution are alike. We have in another paper (this Report, p. 369) pointed out that the globulin of the castor bean shows a similar behavior, and that the part precipitated by saturating with salt is a derivative of the part soluble in saturated salt solutions. We have further shown that the addition to a solution of edestin of a quantity of acetic acid too small to detect after mixing with the proteid, causes a precipitation of the edestin on saturating its solution with brine, and that under these conditions, the proteid otherwise behaves like the globulin from the castor bean and sunflower seed.

As helianthotannic acid contains about 53.0 per cent. of carbon the presence of two per cent. of this acid in our preparation would but slightly raise the figures obtained for carbon and reduce those for nitrogen by about 0.35 per cent. The composition of the purer preparations which we have obtained differ from edestin to about this extent.

It is therefore our opinion that the sunflower seed contains as its principal proteid the globulin edestin, but that as obtained by extraction from the seed, this is mixed with helianthotannic acid, from which we have not succeeded in separating it completely.

Having thus found that a large part of this globulin is insoluble in saturated salt solutions under all the conditions of our tests, we were led to repeat Vines' experiments, but have been unable to confirm his observations, the aleurone grains appearing to be wholly unaffected by saturated salt solution after treatment of the seed with alcohol.

THE PROTEIDS OF THE COW PEA, (Vigna Catjang.) BY THOMAS B. OSBORNE AND GEORGE F. CAMPBELL. The proteids of this plant have never been, to our knowledge, the subject of study. Because of its great and increasing agricultural importance, and as a plant differing botanically from those included in our investigation of "legumin," the proteids of its seeds have much interest. The material examined was prepared by coarsely grinding the peas, separating the black seedcoats by a current of air, and then grinding the coarse meal to a fine flour. Two kilograms of this flour were treated with a quantity of ten per cent. sodium chloride solution, the extract was strained through fine bolting cloth and allowed for three hours to deposit the greater part of the suspended starch. The extract was then run through a DeLaval centrifugal separator, whereby most of the remaining suspended starch and fiber was removed, and lastly was filtered perfectly clear by passing through a thick layer of filter paper pulp. The extract was saturated with ammonium sulphate, the precipitated proteids were collected on a filter, and dissolved in brine. The solution was filtered perfectly clear and dialyzed for four days.

The proteid, thus separated, in the form of spheroids, was designated A, and the solution filtered therefrom was marked B. A was collected on several paper filters. One portion was washed very thoroughly with water and with alcohol and, dried over sulphuric acid, gave preparation 1, which weighed 29.7 grams. The rest of A was dissolved in one liter of five per cent. sodium chloride brine, and the solution filtered perfectly clear. On adding one liter of distilled water a large precipitate, D, separated, which was allowed to settle over night. The fluid, C, was then decanted from the proteid, D, which had formed a coherent deposit. D was dissolved in 150cc of ten per cent. sodium chloride brine with which it readily yielded a clear solution, and water was added to make the volume one liter. A rapidly settling precipitate appeared that soon united to a coherent layer, from which the nearly clear liquid was decanted. After thoroughly washing this precipitate with water and with alcohol and drying it over sulphuric acid there resulted 50.9 grams of preparation 2. The solution decanted from 2 was treated with 200cc of water, causing a precipitate which, washed and dried as before, gave 12.83 grams of preparation 3. Similarly the solution.