286 University of California Puhlications in Botany [Vol. 5 



pass I'rom group A to group B in any series, however, the amount 

 of the other salts diminishes, so that B contains two-thirds, C one-half 

 and D one-fourth the amount of a given salt contained in the pots in 

 group A. This arrangement gives, of primary importance, a series 

 in which the total concentration of nutrient salts varies widely and 

 in addition gives a distinct variation in each individual nutrient both 

 within a group and between different groups. 



The results of these various experiments are almost entirely nega- 

 tive so far as the effort to inhibit flower-fall completely is concerned. 

 It was. however, established that when the total concentration of 

 available nutrients falls below a certain point almost all the young 

 fruits are held (m the plant for a very considerable period of time 

 in place of falling, in the majority of cases, shortly after anthesis, a 

 condition characteristic of the flowers of sister plants under field 

 conditions. An appreciable number of fruits on many plants came 

 almost to maturity before falling whenever the total concentration 

 was low. The effect of low total concentration of nutrients was also 

 seen in connection with pollinations of hybrid stigmas with normally 

 matured parental pollen. In the field it is difficult to make such 

 pollinations at just the proper stage in the development of the hybrid 

 flower and many attempts are necessary to secure even a little seed. 

 This seems to be due to the fact that under field conditions the absciss- 

 layer in the hybrid pedicel is formed when a certain stage of pistil 

 development is reached without pollination and the fertilization of 

 the few good ovules matured. Thus in the field successful pollination 

 could be secured only when the pollen was applied to the stigma ju.st 

 before the opening of the anthers, the latter often taking place before 

 the complete opening of the flower. In the case of the pot cultures 

 of sister plants which were given a low total concentration of nutrients 

 almost every attempted pollination was successful and to a considerable 

 extent the stage of development of the flower was of no .significance. 

 It must be definitely noted, however, that no increase in the number 

 of ovules capable of fertilization accompanied these more generally 

 successful back-crosses. Further evidence in the above connection is 

 given by the fact that plants growing for a number of years in pots 

 of unrenewed garden soil have in successive years shown a longer and 

 longer period of retention of flowers and fruits. 



On the other hand, so far as our experiments are concerned, there 

 is no indication that an increase or decrease in the normal concen- 

 tration of anv one constituent of a nutrient solution will affect flower 



