226 [March, 



In a set of larvae reared carefully the 3 (^ and 4 $ moulters were 

 fully as numerous as tlie 4 (^ and 5 $ moulters, but in a set reared 

 carelessly, in so far that many were kept in one vessel, and their food 

 being taken from different trees (always pear) at different times, was 

 sometimes obviously unpalatable ; there were hardly any 3 (^ and 4 $ 

 moulters, but there was a large mortality among these, and it is equally 

 possible to conclude that the hard conditions of life killed the 3- and 

 4-moulters as that they changed them into 4- and 5-moulters. 



These points then remain for further investigation. What 

 determines whether or no a larva shall moult the additional time ? 

 Is it already pre-determined before it leaves the egg, or does it depend 

 on conditions of weather, feeding, &c. ? "Would it be possible by 

 selection to raise a race of antiqua, consisting entirely of 3- and 4- 

 moulters, or again of 4- and 5-moulters, or of 4-moulter3 only ? Do 

 such species in any way occupy an intermediate position between 

 others moulting on the one hand 4 times, and on the other 5 times as 

 a fixed number ? 



Several species which hibernate in the larval stage are known or 

 believed to vary in the number of moults, and hairy larva) which are 

 said to vary in the number of moults, include many species that 

 hibernate as larvae. It is, therefore, of interest to note that O. antiqua, 

 though not hibernating, is closely related to several species that do so, 

 such as O. fascelina, and the allied arctic species which is believed to 

 pass more than one winter in the larva state. 



If hibernation and variation in moulting are related, it would 

 appear probable that the additional moults are a provision for prolong- 

 ing the duration of the feeding period after an exceptionally bad time 

 during the winter, and that the number of moults in any individual 

 may vary according to its x'equirements determined in this manner, or 

 on the other hand a less number of moults may represent what occurs 

 in many species, viz., a rush to reach the perfect state as early as 

 possible, and so escape hibernation and to become double-brooded. 



This suggests the enquiry whether the summer broods of such 

 species as Selenia lunaria and illunaria, which are smaller than the 

 type, and have been described as distinct species, moult as often as 

 the autumn brood. 



Further investigation thus promises to throw valuable light on 

 several questions concerning variability. 



An isolated observation on Acronycta alni bearing on this matter 

 may be worth mentioning. A few eggs were given me, from which I 

 i-eared 5 larvae ; 4 of my larvae moulted 4 times, but the 5th moulting 



