LEWIS. — POLYCH^TE ANNELIDS. 241 



c. Minute Structure. 



The giant cells are never spherical, their axes being of unequal length 

 (Plate 1, Figs. 1, 2, 3; Plate 2, P'igs. 8, 10, 13). Sections through 

 the long axis of the cell are very unlike those through the short axes. 

 In a section made parallel to the long axis, the nucleus is seen to occupy 

 an eccentric {)Osition (Figs. 1, 2, 3), and is sometimes found lying in a 

 sort of outpocketiug of the cell (Plate 2, Fig. 10). The nucleus, al- 

 though large and distinct, is relatively much smaller than the nuclei of 

 the smaller cells of the nerve cord (Plate 3, Fig. 21), and has a 

 well marked nuclear membrane. It usually contains only one large 

 nucleolus, but sometimes two. The axes of the nucleus are frequently 

 unequal, measurements of a dozen or more nuclei having given a varia- 

 tion from llju to 20(1 for the long diameter, and from 8^ to 12;U for the 

 short diameter. But sometimes the nucleus appears round, or nearly so, 

 when cut in the long axis of the cell, — the direction which shows the 

 inequality of the nuclear axes, if any exist. In sections perpendicular to 

 the long axis of the cell the nucleus occupies a central position. 



Near the nucleus, but nearer the centre of the cell, is a peculiar struc- 

 ture which seems as constantly present as the nucleus itself. In none of 

 the giant cells prepared by either of the methods given was it ever want- 

 insf, although the details of the structure were not altogether uniform. 

 The new species, Clymene producta, was much the more favorable of the 

 two for the study of this structure, and all the giant cells figured are 

 from that worm. No similar body has been mentioned for giant cells in 

 other annelids, although the giant cells themselves have been described 

 in the nervous system of annelids many times. An article by von Len- 

 hossek ('95*), which was received shortly after I first discovered these 

 cells, describes for the nerve cells of the frog, under the name centrosome 

 and sphere, a structure which resembled so strongly the peculiar body in the 

 giant nerve cells of Clymene producta, that there could be no doubt of 

 their being identical. 



For the present, I shall designate the entire structure by the name 

 "sphere," and later discuss the application of the term. 



In the case of these giant nerve cells there could be made out in the 

 sphere the following well marked regions. (Plate 1, Figs. 2, 3; Plate 

 2, Figs. 11, 12 ; Plate 4. Fig^. 24, 26 ; Plate 5, Figs. 29-32.) 



(1) The outer part of the differentiated region consists of a broad zone 

 of rather coarse granules ; (2) within this is a smaller central area of 

 nearly homogeneous protoplasm ; and (3) within this central area a 



VOL. XXXIII. — IG 



