L. DONCASTBR 11 



at which the various stages of the polar divisions take place, for the 

 usual method in fixing the eggs was to determine as accurately as 

 possible when the moth began to lay, and about two and a half hours 

 later to fix all the eggs laid. In eggs fixed less than an hour after 

 laying, usually the first polar spindle is found ; in eggs fixed two and a 

 half hours after laying began, the latest stage found is commonly the 

 anaphase of the second division. The equatorial plate stage of the second 

 division appears to last for some time, since it is the stage most fre- 

 (juently found. This is fortunate, for it is the only stage at which 

 the chromosomes in the inner and outer spindles can be counted with 

 certainty. Even then, on the average not more than one egg in ten 

 has the chromosome groups cut in a plane in which they can be 

 counted, so that of nearly 600 eggs cut and examined, only about 

 60 gave at all reliable counts. The proportion is considerably higher 

 in batches of eggs which could be shelled easily without damaging 

 them, so that accurate orientation is possible ; in such batches, as many 

 as one in three or four may give good iigures. 



During the first polar division, a mass of granules which stain 

 deeplj' with iron haematoxylin is left in the equatorial plate as the 

 chromosomes travel to the poles (Fig. 14). This remains in the second 

 division as a flat plate of fine granules between the outer and inner 

 mitotic figures. When, as happens sometimes, one chromosome lags 

 behind the others, it can easily be determined to which spindle it 

 belongs, even when it is in the next section to the remainder of the 

 chromosomes, by its position on the outer or inner side of the granule 

 layer. In most cases, however, all the chromosomes of a group (outer 

 or inner equatorial plate) are found in one section, and except where 

 the contrary is stated, all the figures have been drawn from such cases. 



One abnormality which was observed must be mentioned, since it 

 may have considerable theoretical importance. In one egg of brood 

 '13.4, and in two of brood '13.42, there are two separate polar division- 

 figures, near together at the anterior end of the egg (Figs. 15, 16 a, b). 

 In the egg of '13.4 and in one of those of '13.42 each figure is in the 

 equatorial plate stage of the second division, with an inner and outer 

 spindle; in the second egg of '13.42, each figure shows three vesicular 

 polar nuclei, and two female pronuclei each conjugating with a sperm- 

 nucleus (Fig. 16 «, b). There must therefore have been two nuclei in 

 these eggs, which are undergoing maturation simultaneously. If in such 

 an egg one nucleus underwent maturation in such a way as to give rise 

 to a male-producing pronucleus, the other to a female-producing, a 



