194 THE entomologist's RECORD. 



photographs by Mr. Tonge (all the others are by Mr. F. N. Clark). 

 The larvae are on a spray of ('dtis austyalis. I need say nothing about 

 the pupa, as I described it in F.nt. ilec, vol. xii., p. 284 (pi. xv., figs. 7 

 and 8). 



It is to be noted that the larva retains a cryptic coloration like the 

 Satyrids, and has not developed a warning aspect like the Vanessids. 

 The subsegmentation into four equal subsegments agrees with the 

 simpler instances in Pierines and Satyrids, these usually have more 

 subsegments, those that have four are probably more primitive in this 

 respect, though they may not be so in others. 



The hairs retain throughout the appearance at the extremity of 

 having an opening, with some fine points round the margin (pi. xii., 

 fig. 2, hairs of first instarx500), but when the 4th instar is reached 

 the extremity is not enlarged at all, with the trumpet-like extremity of 

 the 1st and 2nd instars, but tapers fairly uniformly to the end, where, 

 however, of course, it is not pointed, but as if truncated or broken oflf, 

 until magnified enough to show the actual structure. As already noted 

 the finer hairs in the last instar and those beneath are of the ordinary 

 pointed form. 



The structure of the larva is chiefly of interest so far as it may 

 throw a little more light on the true affinities of L/ibythea. 



There are several items that may be taken into account. The 

 general appearance of the larva is certainly that of a Pieris or Kvcldoe, 

 this one might pass over as a result of "convergence," in larvae living 

 exposed on green foliage, were it not that the hairs, if not precisely, 

 are very nearly, identical with those of many Pierids, in having an 

 extremity formed for exuding fluid. The gradual change in the 

 number and arrangement of hairs as each instar arrives is much the 

 same as in Fierix, but without ignoring this, its importance is small, 

 as it is much the same in many different larvfe. The circles of hooks 

 on the prolegs are so actually or nearly complete as to suggest that we 

 have in them a record of affinities with much lower groups, no other 

 butterfly that I know has complete circles of hooks, except among the 

 Hexperidae. 



The wing outline of the imago, and the foodplant of the larva, 

 seems to point unmistakeably to alliance with Vam'xsa. The imaginal 

 colouring gives the same sug'gestion. The pupa is Nymphalid, but 

 certainly not Vanessid, perhaps more nearly Danaid. The egg is 

 almost if not quite Vanessid. The palpi of the imago are, I believe, of 

 no phylogenetic significance. Scudder (1889) is unable to suggest any 

 utility they may have. I set a specimen or two to show their real 

 meaning, in agreement with the attitude of living specimens. I 

 called Mr. Tutt's attention to them, and I think they passed into his 

 collection ; Avhere they are now I don't know, but Mr. Tutt published 

 a note on the point in Knt. Ilemnl, vol. xi., p. 239. The cryptic 

 habit, so built up must have proved very useful, as it is still pronounced 

 throughout the whole sub-family. 



The neuration is rather Satyrid than Vanessid. The normal 

 structure of the female fore-tarsi, would certainly place Libytltea at the 

 base of the Nymphalids. In this as in other cases we must remember 

 that lAhi/t/wa has had as long a time to be modified in various 

 directions as its relatives, so that whilst such primitive characters as 

 it possesses are probably significant, specialised characters, whether 



