External Conditions and Manuriiig. 



317 



way I filled two beds of about two square meters each 

 in the summer of 1892. They gave a very characteristic 

 half curve when the inflorescences were plotted according 

 to the number of flowers on that part of the axis which 

 had proliferated. These numbers varied from to 10 

 and were distributed as follows (August 1892) : 



NUMBER OF FLOWERS ON THE PROLONGATION 



OF THE AXIS 



Inflorescences 0123456789 



On the 1st bed 325 83 66 51 36 36 18 7 6 1 



" " 2nd bed 403 97 62 35 46 20 20 14 11 3 



Fig. 6o. Trifolium repens perumhellatum. Four different 

 inflorescences from the same culture, 1891. A, with thick 

 continuation of the main axis ; B, the continuation thin 

 and sparsely covered with flowers ; C and D, with three- 

 fold continuation. 



We see that one-half of the inflorescences on this 

 plant had proliferated this time, partly as the result of 

 selection, partly on account of improved cultivation. Most 



