242 PROCEEDINGS OF THE AMERICAN ACADEMY. 



minute deeply staining body, or sometimes three or four such bodies. 

 This corpuscle or group of corpuscles I call centrosome, employing von 

 Lenhossek's term, although in the structure so designated there is a lack 

 of agreement with von Lenhossek's centrosome, which will be pointed out 

 later. There are radiations, which sometimes extend out beyond the 

 sphere into the undifferentiated portions of the cell, but sometimes only 

 through the coarsely granular zone of the sphere. In some preparations 

 (Plate 1, Fig. 2), these radiations traverse both zones, and seem to take 

 their origin from the central corpuscle. In others (Plate 4, Fig. 26), 

 the radiations traverse only the outer more coarsely granular zone, and 

 do not cross the inner more homogeneous area. The rays were usually 

 present in rather small numbers, so that they could be readily counted. 

 In such cases, they were generally separated by nearly uniform intervals, 

 although often they were interrupted over an arc of many degrees. 

 Occasionally the radiations were too numerous to be readily counted. 

 In several preparations, they were so clear and strong that they could be 

 seen without the use of anything higher than an E (Zeiss) objective, but 

 in others they were brought out only by the use of an immersion lens. 

 A ring of microsomes appeared in one preparation at a uniform distance 

 from the central corpuscle (Plate 2, Fig. 11). 



The sphere was sometimes rather sharply marked off from the remain- 

 ing protoplasm of the cell, but usually graduated into it. 



In all the specimens of Clymene producta studied, there were found 

 in the regions of the sub-oesophageal ganglion, and nowhere else, giant 

 cells which contained each two spheres (Plate 2, Figs. 8, 9, 13, 14). It 

 was sometimes possible to make out in these cells a central corpuscle and 

 radiations, but not with the same clearness as in many of the cells con- 

 taining a single sphere. In most instances of cells containing two spheres, 

 the spheres are very sharply marked off from the rest of the cell proto- 

 plasm, sometimes by a ring of granules, sometimes by a clear space around 

 the sphere. It seems to me a peculiar fact that cells containing two 

 spheres should always be found in this region and not in other parts of 

 the worm ; but I have no explanation of this fact to offer. 



Outside the sphere the contents of these giant cells consist of a clear 

 protoplasm, which appears identical with that of the cell-process and 

 Leydig's fibre. In many cells there is at a little distance from the 

 sphere a band of denser protoplasm, which partially surrounds the sphere, 

 but is interrupted on the side toward the nucleus, apparently owing to 

 the presence of tbiit structure. 



Yjj the vom Rath method, as well as by the use of iron hsematoxylin, 



