16 BULLETIN 120, UNITED STATES NATIONAL MUSEUM. 



■what larger than the achromatic granules of the nucleus, possibly be- 

 cause of their darker staining. These are also connected by fibrillae 

 best seen during mitosis, when they form a longitudinal spindle of 

 somewhat branching fibers lying in a layer within the outer spindle. 

 In the reticulate nucleus, these chromatin granules are scattered, show- 

 ing no discernible special arrangement. Their behavior has not yet 

 been followed through the whole of mitosis. During the anaphase 

 the picture is clearest (see Protooyalina saturnalis, fig. 37, /, p. 63). 

 Then these granules are aggregated into linear groups lying along 

 the longitudinal fibers of the inner spindle. These groups differ 

 from one another in length and in number of component granules. 

 Protoopalina intestinalis has not yet been carefully studied as to 

 the detailed character of its granule groups, but in Zelleriella 

 [" Opalma^^] aritlUiensis there are 10 of these lines of granules, 

 differing from one another in length and in number of component 

 granules. Zelleriella antilliensis has 10 of the large chromatin masses, 

 so that the number of the chromatin masses and that of the groups 

 of granules are the same. Eeexamination of my slides of Protoopalina 

 intestinalis and P. caudata shows about eight linear groups of gran- 

 ules in the former and about six in the latter, these species shoAving 

 respectively eight and six chromatin masses. My preparations of 

 these species are, however, not now in quite good enough condition 

 for one to say definitely that their groups of chromatin granules are 

 respectively eight and six. These are the counts obtained, but the 

 slides, now over fourteen years old, are not clear enough to warrant 

 a definite statement. I have not had enough favorable material of 

 Zelleriella antilliensis to allow detailed study of many nuclei and I 

 cannot say if the groups of granules show constant and characteristic 

 differences in the number of their component granules. I hope to 

 obtain material for testing this point and determining the question 

 of individuality in these granule groups. 



The granule groups gather in the equator of the nucleus at the 

 mitotic phase of the mitosis and then divide each group into two, one 

 daughter group of each pair migrating to each pole of the nucleus. 

 No observations have as yet been made as to longitudinal splitting 

 of the groups of granules. 



We have referred to chromatin masses and groups of granules. It 

 seems, however, justifiable to call them massive chromosomes, or 

 macrochromosomes. and granular chromosomes, or microchromo- 

 somes, and hereafter in this paper we will so name them. 



The nucleolus does not disappear or divide during mitosis, at least 

 it does not generally do so.^- It passes bodily into one of the two 

 daughter nuclei, and in the other daughter nucleus a new nucleolus 



'* I have novel" seen evidence of the division of a nucleolus. 



