Sex-Determination in the Gall- Fly, N. lenticularis. 191 



which it is not easy to place in either group with complete confidence. For 

 example, the eggs shown in Series VI, 1-7 (Plate 7), are intermediate in 

 their early stages between the top-shaped and spherical nuclei, and the later 

 stages, though not very clear, might belong to either type. Similarly, while 

 the early stages of the series shown in Figs. I, 1, 2, are clearly of the 

 top-shaped type, the later figures (I, 8 and 4) might well belong to the 

 second type. While, therefore, there are some indications of two distinct 

 types of maturation process in the eggs of different females, the results 

 obtained do not make these differences so certain as to justify any confident 

 conclusion that they correspond with the male-producing and female- 

 producing offspring. 



Whether the differences noticed in the eggs laid by different females are 

 connected with the fact that the flies to which they give rise are either 

 male-producing or female-producing may be doubted, but in any case the 

 nature of the nuclear divisions during maturation is so remarkable as to 

 need a somewhat fuller description. After the nucleus has swollen up, 

 whether it be of the " top-shaped " or " spherical " type, its membrane 

 disappears, usually first at the inner pole (Fig. VIII, 1), and from the 

 nuclear reticulum fine strands are drawn out towards the centre of the egg. 

 Some figures suggest that these are at first loops, but others {e.g., Figs. VIII, 1 ; 

 IX, 1-6) that they are strands of chromatin not connected with each other 

 at their inner ends. Many figures show that these strands are double at 

 their bases (IV, 3; VII, 4, 5, 6; IX, 1, 2, 3, 4), and when, as not 

 infrequently happens, they appear to arise singly from small masses of 

 chromatin (III, 4 ; IV, 2 ; VII, 2), it is possible that these masses are 

 formed by coalescence of two threads in consequence of somewhat defective 

 preservation. In good figures it is usually clear that these strands are 

 about 10 in number, and when only a smaller number is visible it is possibly 

 because some are longer than others, and only the long ones are sufficiently 

 clear of the reticulum to be recognised. While these strands are being 

 formed at the inner side of the nucleus similar but shorter rods or loops are 

 formed also at the opposite side, towards the edge of the egg, and the two 

 series are for a time connected by a network, on which the chromatin is 

 sometimes aggregated into deeply -stained masses. 



The subsequent behaviour is extremely hard to follow, and is possibly 

 different in the two classes of eggs, the existence of which has been suggested 

 above. What appears to happen is that the reticulum between the inner 

 and outer rods disappears, perhaps by becoming concentrated into small 

 chromatin masses attached to the proximal ends of the strands (Figs. Ill, 4 ; 

 IV, 4), and the inner group of strands then separates itself from the outer 



VOL. lxxxix. — 15. [t 



