242 Hyhriih of Blxtoiune Moths 



I allowed one larva to pupate ; it was fortunately a male and the testis 

 was preserved about two weeks after pupation. 



The single larval ovary is most unfortunately not very well pre- 

 served. It contains only one oogonial equatorial plate which is 

 sufficiently in face for the chromosomes to be seen at all clearly, and it 

 is not good enough to provide an accurate count of the chromosomes. 

 It shows, however, a mixture of large and small chromosomes such as 

 I have described in the spermatogonia! plates of the converse cross. 

 Other odgonial divisi<ins, which lie more obli(|uely, also show chromo- 

 somes of two very different sizes. Owing to the defective preservation, 

 the development of the oocytes cannot be made out completely. There 

 is a stage with a thin spii'eine thread massed at one side as in the 

 normal early synapsis stage, and lower down the tube there are larger 

 nuclei with a thicker thread, but I have not found any nuclei in which 

 the thread has segmented into sejiarate chromosomes, such as are found 

 in normal older oocytes. The ovary, however, is so small that this 

 might perhaps not have occurred at this stage of development even if 

 normal synapsis had taken place. Since the development of the 

 chromosomes is much more clearly seen in oocytes than in spermato- 

 cytes of Lepidoptera, it is disappointing that in my only ovary the 

 stages required are not represented. 



Spennatoffenesis. The testis of the single male pupa is large and 

 well develojjed, and contains all stages up to the spermatids, but no 

 nearly mature spermatozoa such as would probably occur in either of 

 the parent species at the same date. The spermatogonia! divisions are 

 like those of the converse cross ; I have not been able to make counts 

 which I can regard as completely accurate, but in the equatorial plate 

 from which fig. 27 is taken the chromosomes are clear enough to give 

 some approach to accuracy. I count in tliis plate 14 large and 55 small ; 

 in other plates I have counted only 11 or 12 large, and it is probable 

 that two or three reckoned as large in the group figured are zonaria 

 rather than hirtaria chromosomes. It is also possible that two or three 

 small ones counted as two are really single ones in division. The 

 number is sufficiently near to the theoretical expectation of 70 to make 

 it almost certain that complete sets from each parent are present. 



The early stages of the development of the spermatocytes show 

 quite clearly cells with a fine spireme contracted to one side of the 

 nucleus ; it does not differ conspicuously from the corresponding 

 " contraction-phase " of synapsis in the pure species (fig. 28). After 

 this stage the nuclei enlarge considerably, and the thread takes the 



