STUDIES ON SOjrn i:XTI{.\NlJITrAL NKCTAItlHS. 17 



aud as the plant grows proceeds to the upper ones. Sunlight 

 seems to be of the greatest importance to the growth of the 

 nectaries, but not, in most cases, to the secretion of the nectar. 

 d. Phenomena of secretion. 



Out of a large number of extranuptial nectaries, I found only a 

 few that were actually secreting nectar. Borne of these nectaries 

 were of a very small size, especially in the cases of Biosjn/ros Kakiy 

 Callicarpa japonica and Paulownia tomentosa. In these plants, it 

 is difficult to discern any nectar drops on the nectaries. I could 

 infer the actual occurrence of secretion only from the presence of 

 ants on these nectaries. In Ligustrum ciliatum, L. medium and 

 L. japonicum, the nectaries decay, after the secretion is finished, 

 and are covered with a corky layer which is developed from the 

 basal cells of the nectaries. At the next season, there are formed 

 sometimes on the same leaf new nectaries, which then secrete the 

 nectar. The nectaries of P^^unus Laurocerasus retain vigorous activi- 

 ty, producing under favourable conditions fresh nectar repeatedly. 

 When a stem or a branch with extranuptial nectaries on it is 

 cut off and its end is put in water, the nectaries soon begin to se- 

 crete. In a quarter of an hour, small drops of nectar may be 

 observed gathering on the surface of the nectaries. The secretion 

 is a transparent colourless liquid, which has a very sweet 

 taste. It reduces Feh ling's solution, but it gives no reaction of 

 fat or albumen. 



IV. Conclusion. 



1. There are two forms of extranuptial nectaries. One is 

 rej)resented by Polygonum sachalinense and the other 

 by Prunus yedoensis. 



2. The first type of extranuptial nectaries develops from 



