454 BOTANICAL GAZETTE [june 



Material and methods 



The material was collected along the steep, damp, shady banks 

 of Clear Creek, 6 miles south of Bloomington, Indiana. Collec- 

 tions were made 4 days and 5 days apart, from June 30 to July 27, 

 and while some of the anthers at the earlier date failed to show 

 differentiation of sporogenous . tissue, many from the later collec- 

 tions were in the shedding conditions. Often all stages from rest- 

 ing to tetrad would be found in a single umbel. 



Strong chromo-acetic acid was used for killing and fixing, and 

 allowed to act 36 hours, the fluid being changed once or twice 

 during that period, after which the material was washed 18-24 

 hours by means of repeated changes of water, slowly dehydrated, 

 cleared in chloroform, and then imbedded in 52 paraffin. 



Sections were cut 5-12 fi thick, varying with the stages sought. 

 For the spindle and in some instances for the spirem, Flemming's 

 modified triple stain was found to be preferable, although for all 

 other phases, especially the early prophases, Haidenhain's iron- 

 alum-haematoxylin gave the best results, lichtgrun in clove oil 

 occasionally being used for a counter-stain. 



Description 



PRESYNAPTIC AND SYNAPTIC STAGES 



An attempt to determine the structure of the resting nucleus 

 and the origin of the various stages that follow by first studying 

 that resting condition would be difficult and indefinite. The 

 investigation must begin with a stage concerning which there is 

 little dispute, and from this may be traced the subsequent steps. 

 Such a condition is to be found at late telophase of the last division 

 in the sporogenous tissue, even though the chromosomes are more 

 or less united by anastomoses, caused by considerable enlargement 

 of the nucleus which followed the close association at early telo- 

 phase (fig. 1). While still retaining this distinct individuality, 

 although united with one another by anastomoses and at the same 

 time slowly approximating end to end, a series of vacuoles appear 

 along the median longitudinal portion of each chromosome (figs. 

 1-4), this being first made apparent by chromatin staining fainter 



