318 Xutrition and Selection. 



of the perumbellate inflorescences appeared in July. Be- 

 fore July there were 21%, in July 47%, in August 38%, 

 and tliis last record was made on over 500 inflorescences. 



I contini^ed the experiment by planting out some of the 

 creeping stems of this plant into two other beds, of which 

 one consisted of good garden soil, the other of dry sandy 

 soil. In this year under less favorable conditions of 

 growth, the proportion of anomalies was less . Amongst 

 about 300 inflorescences in each bed 6% had elongated 

 on the sand bed and 12% on the control bed. On the 

 former the plants were small, their leaves being smaller 

 and paler and less numerous than the normal. 



This experiment shows that the proportion of anom- 

 alous inflorescences is to a very large extent determined 

 by external conditions even when the method of propa- 

 gation is purely vegetative. 



In the literature of this subject we sometimes come 

 across the view that anomalies are favored by improved 

 external conditions because more nutriment is necessary 

 for their development, and the fasciations, and the multi- 

 plication of the number of leaves, leaflets, etc., are given 

 as instances. As a matter of fact even when the anom- 

 aly consists in a reduction of parts the same relation ob- 

 tains. This is shown by the second instance that I shall 

 give. This is Potcntilla anscrina, of which I found some 

 plants with occasional tetramerous flowers near Hilver- 

 sum in 1889 among the ordinary pentamerous ones. I 

 planted them out and divided them, and in the summer of 

 1891 I selected the best specimens and made sure that 

 all the runners were still organically connected. In 1892 

 I planted out one-half of tliis on a manured bed and the 

 other half on an unmanured bed adjacent to it. 



From the middle of July to the beginning of August, 



