148 LOUISE B. WALLACE. 



follows with utmost care the development of the spermatid 

 nucleus, centrosome and idiozome. He, like Wagner, undoubt- 

 edly mistook the compact accessory chromosomes for a nucleolus 

 which is conspicuous and often surrounded by a clear area in the 

 spermatids. According to Bosenberg, this nuclear element dis- 

 appears prior to the formation of the chromatin plate at one side 

 of the spermatid nucleus. As a matter of fact, however, the 

 accessory chromosomes (" nucleolus ") form an important part 

 of the chromatin plate in half of the spermatids, first stretching 

 across the side of the nucleus where the chromatin granules later 

 accumulate. Bosenberg traced the subdivision of the centrosome 

 into proximal and distal portions. The distal centrosome 

 migrates to the cell-periphery and from it grows out the delicate 

 extracellular axial filament. The proximal centrosome moves 

 over or through a portion of the nucleus and later becomes com- 

 paratively large and pear-shaped. It is then regarded as the 

 connecting-piece, or middle-piece, the latter term being thought 

 inappropriate in the spider spermatozoon. The apical body is 

 derived from the idiozome vesicle and contains a filament which 

 bears a small granule. The latter is derived from the connect- 

 ing piece. My observations, so far as they have gone, indicate 

 a close agreement between the transformation of the spermatid 

 of Lycosa and that of Agalena. 



To Berry ('06) belongs the credit of first reporting dimorphism 

 of the spider spermatozoa although in her brief paper on Epeira 

 she was not able to bring forward much data in support of this 

 view. In the telophase of the last spermatogonial division, one 

 chromosome appears to have no mate and is therefore regarded 

 as the odd chromosome which persists as a single, univalent 

 element in the rest stages, becoming longitudinally split in the 

 spireme. In the first maturation mitosis, the odd chromosome 

 is carried to but one pole and while it was not identified in the 

 second maturation mitosis it is thought to divide along the line 

 of the original longitudinal split, the resultant halves being car- 

 ried to the opposite poles of the spindle. This view is supported 

 by the fact that apparently one half of the spermatids contain a 

 single chromatin mass while the other half do not. 



Knowing by experience the difficulty of accurately counting 



