40 



Chloride method, E. B, Wilson ('Ol) found two widely different types 

 of chromosome formation. The two types, however, did not co-exist 

 in the same series of eggs. In one type the nucleolus resolved itself 

 into a spireme which broke up into chromosomes; in the other type 

 the chromosomes arose from the nuclear reticulum. 



If the chromosomes actually come from the nucleolus at matu- 

 ration, the oocyte history in favorable cases ought to reveal their 

 entrance into this structure. I have accordingly undertaken to trace 

 the development of the egg of Asterias forbesii from the last series 

 of oögonia through the entire growth period and thus follow the 

 history of the chromosomes from their first origin from the spireme 

 following the contraction phase (synezesis — McClung — '05) to the 

 time of their reception into the first polar spindle. My results show 

 conclusively that in this Echinoderm form, at least, the chromosomes 

 do not arise from out of the nucleolus. Frequently the chromosome 

 group is in such close superficial contact with the nucleolus as to give 

 the appearance of a nucleolar origin. 



As is now well known, when the immature eggs of starfish are 

 shaken from the ovary into sea water, they immediately begin to form 

 the polar bodies. In ovaries opened during July and August about 

 70 7o of the growing oocytes had attained to a stage of development 

 when they could be thus induced to mature. These were preserved 

 at intervals of 10 min. through a period of two hours. The sea water 

 was kept by frequent changes as near as possible within the limits of 

 normal ocean temperature. Of every series, eggs were allowed to 

 develop to the early segmentation stages, and only such series were 

 saved as showed at least 40 7o of segmenting ova among the residue. 

 Thus one may be reasonably certain that at the lowest estimate about 

 one half of the maturing eggs previously preserved show normal stages 

 in the developmental process. 



Both the ovarian material and the free ova were best fixed in 

 sublimate-acetic. The most serviceable staining combination is Heiden- 

 hain's Iron Hematoxylin followed by orange G. counterstain. The 

 sections were cut at 7 microns. Four complete series of eggs have 

 been studied and give results in accord with each other. The ovarian 

 material (collected during July, August and December) also shows 

 identical conditions in many ovaries studied. 



The oögonial cell of the last order has a nucleus of 2.5 microns 

 dia. It contains an intensely chromatic nucleolus of about one third 

 this dia. enmeshed in a very delicate and wholly achromatic linin 

 network. The chromatin content of the nucleus increases steadily in 



