266 THE entomologist's record. 



in that species, and were in continuation of those on Orgyia antiqiia^ 

 presented to the Woolhope Chib in 1882, and published in the E. M. M. 

 in March, 1887, and which elicited a valuable note from Prof. Riley 

 in the May number. I was so far fortunate that I have not only 

 noted variations in the number of moults, but have found also 

 variation both in plumage and habits to be associated with the variation 

 in moulting, and in such a way as to give some hints as to the meaning 

 underlying them. The subject is one that does not appear to have 

 attracted much attention, and I do not recollect seeing any record of 

 observations exactly similar to those I have made on A. cai'a. 



One reason why so little has been done in this investigation is, no 

 doubt, its tediousness and inherent difficulty. The closest attention is 

 necessary to make sure of the moulting of any larva. It was the 

 custom of Mr. Hellins, in order to secure an exact record of anj' 

 particular species, to isolate an individual larva for sjiecial observation, 

 but in this research we must have more than this, \ve must have an 

 exact record of the moultings, &c. not of one larva only but of 50 or 

 100, or, if jDOssible, of an entire brood ; and so much has the difficulty of 

 this been felt, that some of our best oljservers cannot say what is the 

 number of moults in species of which they have reared large numbers, 

 still less what, if any, variation occurs in connection therewith. I have, 

 however, devised a tolerably simple method of achieving this result, 

 and should hope to see the subject worked out in detail with a number 

 of species, when it is applied by some of our enthusiastic students of 

 larvfe. Until the ineans of ol)servation are greater, such hypotheses as 

 we may be tempted to form in exjjlanation of the observed facts must be 

 of a very tentative nature. My method is simjdy as follows : — As soon, 

 as some larva? lay themselves up for moulting the remainder are 

 removed to another jar, which becomes jar 1, the original being now 

 jar 2, and as soon as any in jar 2 have moulted they are moved on to 

 jar 3, and so jar 1 contains larvai in first skin, feeding ; jar 2, in first 

 skin, laid up for moult ; jar 3, in second skin, feeding ; jar 4, in second 

 skin, laid up for moult, and so on. There may, of course, especially as 

 the larva? gi-ow, or as varieties of habit appear, be several jars of the 

 same number — thus jar 7, feeding, in fourth skin, may be rejieated 

 several times, or be jar 7, jar 7a, jar 7b, according to variation. 



It may be well to begin by describing the larva and larval history 

 of what we may take to be normal caia under ordinary circumstances. 



The eggs are laid in regular hexagonal order in batches of 100 to 

 300 on any flat surface, the underside of a leaf being the usual site, and 

 like the " ermines " and many Noctv^ that la}' their eggs in a similar 

 way, the leaf of a tree or shrub many feet from the ground seems to be 

 often selected, the young larva?, of course, dropping to the ground 

 amongst the low plants, on which they feed. The eggs are nearly 

 sjiherical, 1-00 mm. in diameter, creamy or, quite at first, greenish- 

 white, changing colour to an orange creamy in a da}' or two, but not 

 further till shortly before hatching, when they get very dark chocolate 

 and then black. They have a pearly lustre, and a very delicate trans- 

 parent appearance, but this is deceptive, as they are tolerabty hard and 

 solid, and none of their contents can be seen at all, though their 

 mirror-like surface reflects surrounding objects, and for some time led 

 me to think I was observing some interior structure, although I could 

 not make out what. The surface is mapped out by a very fine network 

 of raised lines, usually in hexagons. 



