240 ' Thomas H. Montgomkry, Jr.. 



into the ovaiial tubes and not, as earlier observers had thought, into 

 a cavity outside of the ovaries. In Thcridium the greater part of 

 the wall of the ovaries is composed of a high, cylindrical, glandular 

 epithelium, and only the deeper basal cells become functional oocytes ; 

 the stalk attaching each of the latter is made up of a few modified 

 cells which send around the oocyte a delicate but apparently quite 

 complete envelope, so that each follicle seems to be quite closed. 

 The haeraolymph bathes these follicles and is probably the main 

 source of nourishment for the oocytes. 



It has not been my intention to describe the structural changes 

 of the early oocytes, for this has been done by various preceding 

 writers, notably by van Bambeke (1898), But I would call atten- 

 tion to the fact that in the smallest oocytes there is a plasmosome 

 distinct from a chromatin reticulum, but that the latter graduall}' 

 concentrates on the surface of the plasmosome, thereby forming a 

 nucleolus composed, as Obst (1899) has shown for other aranead 

 genera, of erythrophilic (plasmosomic) and cyanophilic (chromosomic) 

 substances. When this large nucleolus is established the remainder 

 of the nucleus is occupied by an achromatic mesh work, possibly 

 linin. Contrary to the case in most spiders, there is no yolk nucleus 

 at any stage. 



The latest found stage of ovarian ova showed the nucleus 

 amoeboid and close to the surface of the egg, and the chromosomes 

 concentrating preparatory to the maturation mitoses (Fig. 1, PI. 4). 

 When they first become distinguishable they are in the immediate 

 vicinity of the plasmosome, and not in any other part of the nuc- 

 leus; this shows that they are separating from a plasmosome to 

 which they had previously been apposed, analogous to cases described 

 within recent years for the oocytes of a number of animals. The 

 vacuolated plasmosome breaks into fragments as the chromosomes 

 disengage themselves. Fig. 1 shows the whole nucleus, and Figs. 2a 

 and 2b, from successive sections, illustrate on a much stronger scale 

 of drawing the larger plasmosome, its smaller disintegration pro- 

 ducts, and all the chromatin loops of this nucleus. The latter are 

 very slender and beaded, by their deep stain in contrast with the 

 surrounding achromatic network. The total number of the chro- 

 matin loops shown in Figs. 2a und 2b appears to be 26. but pro- 

 bably the two lettered A, a in Fig. 2a are merely ends of the 

 similarly lettered loops of Fig. 2b, and in that case the total number 

 would be 24, exactly twice that of the chromosomes of the polar 



