198 THE entomologist's record. 



structure, the forewing of 0. ijonostit/iiia is a somewhat shapeless flap, 

 whilst that of (>. ontiqua (pi. vi., fig. 4) still retains something of the 

 true shape of a wing. 



Passing over for the moment <>. ericae and its allied species — 

 though they follow most closely on 0. i/onostitpDa — and taking up 

 0. aiiroliiiihata, we find a further very considerable degeneration. 

 The antennae (pi. iv., fig. 1, and pi. vii., fig. 17, where they are to same 

 scale as those of O. antiqua and 0. (lonoxtiniua are represented) are 

 reduced to an extreme degree, being only one-fourth the length of 

 the antenna of (K i/onoftti;piia, and, though it is still possible to recognise 

 that it is really constructed of some 20 joints, they are all anchyiosed into 

 one piece, except, perhaps, the basal joint may be free, in which case 

 there are two separate pieces. The legs (pi. iv., figs. 2, 3, 4) may be 

 said to be roughly half the length of those of (>. iimioKtii/nia. Their chief 

 feature is the anchylosis of the tibia and all the pieces of the tarsi 

 into one piece in all the legs. The component pieces are still, 

 however, sufficiently marked out to enable one to say that they still 

 constructively exist. They must still necessarily be used in the 

 journey which the moth makes from one compartment of her cocoon 

 to the other. The wings are very much reduced, but still are repre- 

 sented (pi. iv., figs. 5, 6). 



When we come to (>. sjileiulida, we find practically — I may say 

 actually — no trace of antennae and no remains of wings, unless we 

 dignify by these names some doubtful scars marking their position. 

 The legs are much reduced in length, and the tarsi are obsolete (see antea 

 vol. xiv.,pl. i., and pi. v., fig. 16, where it is enlarged 24 times, to be 

 remembered in comparing it with figs, of other species which are 

 enlarged 14 times), rather, perhaps, considering the condition of the 

 anchylosis of tibia and tarsus into one piece in (>. aiirolinibata : what 

 looks like the tibia in (>. splouliila is really the conjoined tibia and 

 tarsus. Though so short and degenerate, they are, nevertheless, 

 comparatively very thick and strong, and provided with robust claws, 

 so that if they are useless for progression, they are very competent for 

 their duty of tearing up the pupa-case and making an opening in the 

 cocoon. 



When we consider the curious habits of (K ericuc, it appears to me 

 that we find them almost as insoluble as those of O. anrolmhata 

 before O. 'inno^tiiiina threw light on them, and we must await the 

 discovery of some species with intermediate habits before we can 

 understand them. One can hardly, however, forbear to speculate. O. 

 ericae is but little more degenerate (specialised) than <>. i/Duostii/ma ; 

 we can hardly, therefore, derive it from (>. atirolinihata. Still, it is 

 very possible that it is so derived if, for (). aurolinibata, we substitute 

 a less specialised ancestor nearer (>. iiojinstit/mo. We have to suppose 

 (). aurolinibata making an external opening at the same time that it 

 opens the second cocoon, and then, on passing into the second cocoon, 

 it is in the position (J. ericae is said to assume. The diaphragm has 

 also to be then omitted. However this may be, the evidence in the 

 case of (>' ;iou(>nti(/i)ia and (>. aurolinibata serves to satisfy us that the 

 changes of habits are made gradually and not in any way per saltiim, 

 and that it must no doubt be the same with O. ericae. One structural 

 item exists to render probable the surmise as to the phylogeny of the 

 habits of O. ericae hazarded above, and that is, that I have noticed, iu 



