282 bulletin: museum of comparative zoology. 



especially since there are no periods of rest, the changes being both 

 continuous and rapid. Until the centriole at the deep end of the first 

 maturation spindle has divided, and the chromosomes of tliat spindle 

 have begun to move apart, there is nothing to indicate that there are 

 two processes, except possibly a difference in the size of the centrosomes. 

 According to the degree of complexity of the changes which accompany 

 this stage, and the length of time required for the metamorphosis, the 

 processes may be designated as the shorter, or direct, and the longer, or 

 indirect. The shorter process is characterized by the fact that the 

 chromosomes remain free and separate from one another in the cytoplasm 

 and by the direct formation of the achromatic figure ; while in the longer 

 process a vesicular nucleus is formed, the chromosomes becoming a i)art 

 of a so-called quiescent nucleus, and the achromatic figure is formed 

 indirectly. The shorter process will be described first. 



(a) Direct or Shorter Process. 



In the preceding section of this paper the changes in the first mitotic 

 figure have been followed to the time of the transverse division of the 

 chromosomes and the formation of the centroplasm and centrioles in the 

 deep aster. The process now under consideration may be said roughly 

 to begin with the anaphase of this stage. Figure 30 (Plate 6) shows 

 a late anaphase of the first maturation figure. The chromatin in tlie 

 polar cell has fused into one 'mass, and the centrosome cannot be 

 seen. The centrosome at the inner pole lies close to the chromosomes, 

 and owing to the obliquity of the section is in fact partly covered 

 by them. The centrosome at this stage may lie so close to the 

 chromosomes that it is impossible to be certain of its outline. In some 

 animals it has been thought that tlie centrosome disappears entirely, 

 — as has been claimed by Byrnes ('99) for Limax, — but in Hamitiea 

 it may always be found when properly stained. The chromosomes are 

 usually arranged in a circle, and in many instances the individual 

 chromosomes are so close together that they seem to be fused ; never- 

 theless, it is probable that they remain separate. The interzonal fila- 

 ments are easily seen extending from one mass of chromatin to the other, 

 and a small " midbody " is recognizable. The medullary layer of the 

 sphere is thin and faintly outlined. A few rays remain attached to the 

 boundary limiting the centroplasm, and within it are to be recognized 

 two centrioles. The evidence of a spindle between the centrioles at this 

 stage is very slight, so that none is represented in the figure. The whole 

 centrosome at this stage is extremely faint and difficult to see. 



