1916] Goodspccd-Ayres: Sterilitij of Nicotiana Hybrids 285 



phoriis was the variant, as was potassium iu group C. In making 

 the applications of the nutrients the constituents of the following 

 nutrient solution were used: 



80 g. sodium nitrate iu 2500 ec. water 



25 g. mono-calcium phosphate in 2500 cc. water 



50 g. potassium sulphate in 2500 cc. water 



20 g. magnesium sulphate iu 2500 cc. water 



At intervals of a week varying amounts of each of the above solutions 

 were added to each pot to give the totals for each nutrient stated in 

 table 3. A trace of ferric chloride was added to each pot. 



The experiments of the first year correspond to the above and 

 the emphasis was similarly laid upon varying the amount of individual 

 nutrients. The range of variation in concentrations and the number 

 of plants employed were, however, considerably less, and a full report 

 of the experiments of the first year would add nothing to that given 

 in table 3. 



The results obtained during the first two years indicated that 

 variations in total concentration of available nutrients, rather than 

 particular concentrations of individual nutrients, might be the im- 

 portant factor in inhibiting flower-fall. To this end the experiment 

 detailed in table 4 was devised to give the wide range of variation 

 in total concentration together with variation in the amount of indi- 

 vidual nutrients. One hundred and twenty seedlings were transplanted 

 to pots containing 2000 g. of washed sand. The salts were all added 

 to the pots along with the sand at the start of the experiment and 

 thereafter the plants were watered at regular intervals with distilled 

 water. 



As shown in table 4, sixty different combinations were secured and 

 a duplicate test, as in the work of the preceding year, was carried 

 through. These combinations were equally divided into three series — 

 I, II and III — and each- series was in turn divided into four groups. 

 In series I, table 4, nitrogen as sodium nitrate was the only varying 

 factor within the group, the amount of each of the other salts being 

 held constant. Correspondingly in series II phosphorus as mono- 

 calcium phosphate was the only variant within the grouj), while in 

 series III potassium ms potassium sulphate was the group variant. 

 Within each of the four groups — A, B, C, aiid 1) of each series — the 

 increase t)!' llic varying nutrient was the same from ])ot 1 to pot 5. 

 Thus in series I, pot 1 in all I'oui- groups contained .02 g. of sodium 

 nitrate, pot 2 contained .2 g. in all foui- groups, and so on. As we 



