On nucleolar structures of the hypodermal cells of the larva of Carpocapsa. 387 



sometimes it does not appear to be so connected bat lies close to the 

 nuclear membrane. In this, as in all later stages, there is only one 

 chromatin nucleolus to a nucleus. In nuclei of a little more advanced 

 stage (Fig. 3) the chromatin nucleolus is larger, and frequently in the 

 form of a hollow sphere with an enclosed eosinophilous globule; it is 

 still attached to a true nucleolus. Finally , the largest and most 

 modified nuclei (Figs. 4, 5), those namely on the distal margin of a 

 foot, are of an elongate or oval form (their long axes perpendicular 

 to the surface of the hypodermis), and often of an irregular amoeboid 

 form; in them the chromatin nucleolus is correspondingly larger, in 

 most cases spherical, either solid or containing small vacuoles, and 

 still opposed to a true nucleolus. For later stages we should have 

 to examine larger larvae as well as imagines. 



The important phenomena in this development are the gradual 

 appearance of a chromatin nucleolus, which at first seems to resemble 

 any chromatin granule, except in regard to its staining intensity, and 

 its opposition to a true nucleolus. This would point to this chromatin 

 nucleolus being a metamorphosed chromatin granule, one which would 

 seem (to judge by its staining reactions) to contain a greater per- 

 centage of pure nucleic acid than do the others. Since at the time 

 when it first appears, i. e. is first recognizable, it is usually situated 

 at some distance from the nuclear membrane, we have a reason for 

 concluding that it is not extranuclear in origin. There is evidence 

 that the true nucleolus undergoes divisions at this period, as shown 

 by the elongate form occasionally seen and the numerical increase of 

 nucleoli ; on this account it might be supposed that the chromatin nuc- 

 leolus is a chemically changed part or whole of a true nucleolus. But it 

 would seem more probable that the chromatin nucleolus is a meta- 

 morphosed chromatin granule which increases in size by the absorption 

 of the substance of the true nucleolus to which it is attached. For 

 at the beginning it shows the pure chromatin stain ; while the 

 eosinophilous globules which subsequently appear in it could be re- 

 garded as portions of true nucleolar substance taken into, but not 

 yet chemically changed by, the chromatin nucleolus. This view seems 

 to best explain its increase in size and concomitant attachment to the 

 true nucleolus; at no stage does it either surround or yet lie within 

 the true nucleolus, so that there is no ground for concluding that it 

 might be formed by the true nucleolus becoming changed chemically 

 from the periphery to the centre, or from the centre to the peri- 

 phery. 



