392 CALKINS. 



membrane has disappeared from that portion of the nucleus be- 

 tween the chromosomes and the central spindle, and that the 

 ends of the chromosomes are connected by distinct fibers — the 

 mantle fibers — with centrosomes inside of the spheres (Figs. 24, 

 26). The chromosomes are then divided longitudinally be- 

 ginning at the ends turned towards the central spindle, and one- 

 half of each chromosome goes to form the daughter nuclei. 

 The nuclei are reconstituted by the subsequent aggregation of 

 the chromatin granules into the large reservoirs while the sphere 

 in each case forms a definite body on the outside of the nucleus. 



The staining reactions of the sphere in Noctiluca are the same 

 as those of the intra-nuclear body in Eitglcna and CJiilonionas, 

 and the same as the cytoplasmic body in Tctraniitus. During 

 mitosis its history is remarkably similar to that of the Ncbcn- 

 k'drpcr as described by Schaudinn ('96) in the case of Parainoeba. 

 I think therefore that there can be no doubt that the sphere in 

 Noctiluca, the Neboikorpcr in Paraniaiha, the cytoplasmic body 

 in Tetramitus, and the intra-nuclear body of Eitglcna, CJiilotno- 

 nas and allied forms are all analogous structures and have the 

 same physiological part to play in the activities of the cell. 



The sphere in Noctiluca however possesses an element dur- 

 ing division which has hitherto not been found in the corre- 

 sponding intra-nuclear or cytoplasmic bodies described above. 

 This element is a distinct centrosome which was first described 

 for Noctilitca by Ishikawa and the presence of which I have 

 demonstrated beyond question. In addition to the centro- 

 somes furthermore there is a second set of fibers — the mantle- 

 fibers — which connect the chromosomes with the centrosomes ; 

 nor have these been found in the simpler nuclei described above. 



A number of the Protozoa agree with Noctiluca in the history 

 of the chromatin, and several observers (Gruber, Hertwig, 

 Brauer) have described the breaking down of large chromatin 

 reservoirs or " nucleoli" as they have been erroneously called. 



When we come to consider the nuclei of the Ciliata and the 

 Suctoria we are met by a new difficulty. The nuclei are di- 

 morphic, and the two forms differ as much in structure as they 

 undoubtedly do in function. I have no new observations to 



