220 THE entomologist's record. 



the tubercles are all red. In the typical larva these red 

 tubercles are so surrounded by red areas as to obscure what 

 is essentially the case, viz., the existence of these lozenges on 

 all segments, only those on 5, 7 and 10 being conspicuous. 

 The tubercles are still large, but the anterior are relatively 

 small and almost in line with the posterior, across the larva. 

 This results in a red band encircling each segment. 



The most remarkable feature of the larva is the disappear- 

 ance of the 2nd segment. This is reduced dorsally to an 

 exceedingly narrow black collar, with two white transverse 

 lines, but is hard to see from the thick fringe of hair passing 

 forwards over the head from segment 3. Laterally, it pos- 

 sesses two tubercles of ordinary type, of a pale cinereous 

 colour. Immediately behind these is the ist spiracle, black 

 and very large. The other spiracles are black, surrounded by 

 paler areas. 



At all stages the head is large, and this gives a somewhat 

 unusual appearance at the moults. The dislocation of the 

 head, which takes place when laid up, has a specially un- 

 healthy and repulsive aspect, the large head giving an appear- 

 ance as if the larva had just moulted, rather than of being 

 about to do so. The progressive atrophy of the 2nd segment 

 adds to the unhealthy look by placing the new head under a 

 very bald membrane, out of place in a hairy larva. 



In the fourth skin, certain larvae become rather larger than 

 others, and moult directly into the last skin, others take an 

 intermediate moult. What I have to say on this point is, 

 perhaps, more interesting than if I had succeeded in making 

 more definite observations. I secured larvae in 1887 from the 

 New Forest, and had several broods from these in 1888. Not 

 being prepared for this variation in moulting, the result was 

 that I concluded I had made some error in counting the moults, 

 and had got muddled. In i88g, however, I found that this 

 variation occurred, about half the larvae adopting each method. 

 I fancied, therefore, that following Orgvia antiqua possibly 

 they were <? and ? . This proved, however, not to be the 

 case ; those that moulted an extra time grew much larger than 

 the others, and when they emerged were larger and finer 

 moths, the wing expanse averaging 39 to 42 mm., against 

 36-37 in the 4-moulters ; but both sexes occurred equally in 

 both sets. In i8go, I intended to make further observations, 

 but most remarkably all the larvae in the brood I especially 

 watched moulted the full number of times (5). In i8gi, I 



