DEPARTMENT OF BOTANICAL RESEARCH. 67 



tration, balance, and reaction, the presence of the solid phase would 

 introduce a factor tending to hold the solutions constant and at the 

 same time to render renewals of solutions less important. 



It has not yet been possible to extend the experiments over the en- 

 tire range of salts which might be used, but the principle has been well 

 established and some of the results are striking. In one or more of the 

 various combinations the following salts have been employed: of cal- 

 cium — carbonate, sulphate, tribasic phosphate, and dibasic phosphate; 

 of magnesium — carbonate, tribasic phosphate, dibasic phosphate, and 

 the double salt with ammonium phosphate; and of iron — phosphate(ic), 

 citrate, oxalate, and the so-called "soluble ferric phosphate" (con- 

 taining citrate). Except in the control culture solutions the only 

 "readily soluble" salts employed were magnesium sulphate and potas- 

 sium nitrate. Without introducing other complexities, there seemed 

 to be no "insoluble" source of nitrate; but magnesium-ammonium 

 phosphate was also used as a possible source of nitrogen. The solu- 

 bility of this salt is but little more than that of magnesium carbonate. 



The methods employed in the preparation of the cultures were the 

 same as in the earlier work, except that relatively large amounts of the 

 "insoluble" salts were used, the quantities being from 0.125 gram to 

 0.5 gram per culture vessel with capacity of 250 c. c, and in the series 

 here refeiTed to no renewal of the cultm^e medium was made. In the 

 case of the two soluble salts, KNO3 and MgS04, small but measured 

 quantities were added at intervals of 10 days or more. Control cul- 

 tures were made with some of the standard solutions of soluble salts, 

 especially the "best" solution of Shive, of mj^ previous work, and of 

 Tottingham, designated respectively solutions A, B, and D. These 

 solutions were renewed as usual. 



The experiments so far completed have been made with the Pacific 

 Bluestem wheat, with a growth interval in the first series of 30 days 

 and in the second of 39 days. In each series one or more of the cul- 

 tures with certain "insoluble" salts gave jdelds higher than the best 

 control, solution B; and in each case at least half a dozen combinations 

 were higher than solution D. The other control, solution A, was un- 

 satisfactory for this variety of wheat and will not be considered in the 

 present discussion. In both series the same combinations of salts are 

 among the highest, but the exact order varies, and this was expected, 

 owing partly to differences in nitrate content. The best culture in the 

 first series, and among the best in the second series, was one containing 

 calcium sulphate, magnesium-ammonium phosphate, "soluble ferric 

 phosphate," and potassium nitrate. A combination with tribasic 

 magnesium phosphate instead of the magnesium salt in the preceding 

 gave values almost as high. Another excellent combination, the best 

 in the second series, consisted of dibasic calcium phosphate and iron 

 citrate combined with small proportions of the soluble salts — mag- 



