No. 513] SHORTER ARTICLES AND CORRESPONDENCE 557 



ferent mean or variability, or both, for the number of ovules 

 which they produce than those which develop into mature fruits? 



Cercis seems to be a form well adapted to an investigation of 

 this question. The ovaries can be quickly dissected out of the 

 flowers which have been dropped into alcohol for preservation. 

 Absolute alcohol renders the walls sufficiently clear that l>y the 

 aid of the light from a properly adjusted mirror of the dissecting 

 microscope the ovules may be counted very easily. Clove oil 

 may be used as a clearing agent, but it soon renders the ovules 

 as well as the walls transparent, and great watchfulness is re- 

 quired in the countings. 



On March 28 a series of one hundred inlloresi-rnces was taken 

 from each of three trees and dropped into alcohol for future 

 study. This was at a time when most of the flowers were open, 

 but when none had fallen from the trees. The whole inflores- 

 cence was taken because (a) it was desirable to avoid any possi- 

 bility of an unconscious selection of flowers in gathering, and 

 (6) because it seemed desirable to determine whether the posi- 

 tion of an ovary on the inflorescence might exert any influence 

 upon the number of ovules produced. 



On April 3 a second lot of material was collected, but in an 

 entirely different manner. Many of the blossoms were ready to 

 fall. The trees were shaken, not too violently, so as to secure the 

 flowers which were ready to drop, but not to dislodge those with 

 ovaries which might continue to develop. In preserving this 

 material all the flowers belonging to inflorescences which had 

 fallen were discarded, since it seemed possible that some of them 

 might have been brushed off by twigs rubbing together. All the 

 flowers were open, and, if their very frayed condition may be 

 taken as a criterion, most of them had been visited by insects. 

 The ovaries were dissected out of the flowers at once and dropped 

 into absolute alcohol. Unfortunately the second collection from 

 Tree II was lost. 



On April 6 a third collection was made in the same manner as 

 the second. 



My intention had been to collect the matured pods of these 

 trees in the fall for comparison with the ovaries taken in earlier 

 stages of development, but, unfortunately, the severe frosts 

 which followed the early opening of spring in St. Louis, in 1907, 

 killed most of the Cercis fruits and precluded the carrying out 

 of this part of the work. 



