A pair of capillary pipettes is made and calibrated to hold one unit, the unit being arbitrarily chosen, in this case a drop of mercury weighing 0.0108 g. After sterilization these tubes are ready for use. A pure culture of Fleischmann yeast is maintained on an agar slant, a transplant being made to fresh agar 48 hours before beginning the test. At the end of 48 hours as small a portion as can be taken up on the tip of a needle is transferred to 10 cc. of sterile Nageli solution (100 c. distilled water, 10 g. dextrose, 1 g. ammonium nitrate, 0.05 g. potassium acid phosphate, 0.25 g. magnesium sulphate) and the tube then shaken for two or three hours in a mechanical shaker. This process does not give absolutely uniform suspensions and occasional clumps are found, but these can be removed for the most part by slow centrifuging and the method has been found satisfactory for results if controlled by a sufficient number of tests. The uniformity of the suspension is tested before use by drawing up with the pipette 5-10 units of the suspension and blowing them out on a glass slide where they are fixed, stained, and the cells counted.

When a fairly uniform suspension is obtained one unit of yeast solution is drawn up into the pipette and one unit of sterile vitamin solution. These are mixed by manipulating the bulb, the pipette is sealed at both ends, and the sealed tube incubated for a suitable interval of time, usually 20 hours at 35° C. At the end of the time the tips are broken at each end of the pipette, a bulb is placed on the large end, and the contents are blown out on a slide, fixed and stained. For a counting slide, Eddy and Stevenson use an ordinary microscope slide on which 5mm. squares are etched. The size holds the contents of a pipette, permits ease of operation when counting with the mechanical stage, and allows the content of ten to twelve pipettes to be placed on one slide. For control, another series of pipettes is prepared and filled by drawing up a unit of the yeast suspension without the unit of vitamin. These are incubated and counted in the same manner.

58 Williams, J. Biol. Chem. 46,113, 1021.

59Eddy and Stevenson, J. Biol. Chem. 43, 295, 1020.

The method was tested with small quantities of Funk's antineuritic vitamin as prepared by him, and the results seemed to strengthen the conviction that this is the responsive causative agent in stimulating the growth of the yeast cells.

Experiments with orange juice showed that the cause of the stimulus is nearly quantitatively removed from orange juice and from navy bean extract by shaking these extracts with Lloyd's reagent. Since Seidell has shown that this reagent removes the curative factor for polyneuritis these results would seem to add to the other evidence that the test is specific for the B vitamine.

Tests with guinea pigs showed that the treatment with Lloyd's regeant did not remove C, in confirmation of the results of Harden and Zilva,60 therefore it is assumed that the C is not responsible in any way for the yeast stimulation.

In estimating the vitamin content of foodstuffs by this method, the probable range of variation in the suspension used was first determined then five to ten tests were made on the substance under examination, and from these all zero readings were eliminated and the rest averaged. In doubtful cases this was repeated as often as necessary.

Funk and Dubin61 have devised a simplified yeast test which they believe is valuable for determining the approximate vitamin activity of a given food substance, although the presence of inhibiting substances may interfere with the reaction to such an extent that the method cannot be used for comparison of the vitamin content of different substances. They prepare yeast suspension by shaking a loopful of a forty-eight hours' pure culture of yeast in 100 cc. of Nageli solution on a shaking machine for three hours. Measured quantities of this suspension are then incubated for twenty hours at 30° in the presence of the unknown vitamin solution, control tubes with a vitamin solution of measured strength being employed. After incubation, the fermentation is stopped by heating the contents of the tubes at 75° for a few minutes, and the amount of yeast is estimated by centrifugalizing in a specially calibrated tube. Using autolyzed yeast as a standard they plotted a curve by means of which the vitamin content of an unknown substance can be expressed in terms of the standard.

The method is said to be sensitive to 0.0001 cc. of autolyzed yeast, but the test is best arranged so that the amounts of unknown substance used correspond to 0.05 cc. of autolyzed yeast.

60 Harden and Zilva, Bioch. J. xii, 93,1018.

61 Funk and Dubin, J. Biol. Chem. 44, 487,1020.

Sousa and McCollum62 also introduced certain modifications of Williams' early procedure, but their study led them to conclude that the method was of little value to a test for the presence of the anti-neuritic factor. Instead of a pen for distributing the drops, they used a statuuner syringe needle having an opening at right angles to its axis, the needle being attached to a piece of glass tubing by a short piece of rubber tubing. It was found that a fairly uniform distribution of cells could be obtained in this manner.

In, order to avoid the difficulty caused by lack of uniformity in; development of yeast cells a Levy blood-counting chamber was used instead of the cover slips used by Williams. Comparison of the two methods showed that a more even and reliable result can be obtained by the use of this procedure. They tested the growth-stimulating effect of aqueous extract of wheat germ made by boiling the germ with water. The results showed clearly that there is a stimulating effect of the extracts on the rate of multiplication of yeast and that in general the greater the amount of extract added the more pronounced the acceleration. Similar results were obtained with the use of alcoholic extract of wheat germ.