202 [August 



others, which are subequal." {Proc. Acad. Nat. Sci. Philad., Sept. 

 1862, p. 378.) This is not known to be the case, so far as I am aware, 

 in any other Ephemerinous imago. We can understand now why it 

 should be so here. In the pupa this joint is abnormally developed to 

 receive the tip of the cai-apace, and therefore, as is not unfrequently 

 the case, traces of the same arrangement are found in the imago, though 

 there are no longer the same special functions performed by the part. 

 There is a feature, too, in the ornamentation of its legs which indicates 

 that it belongs to Ephemerina, and not to Odonata as I had myself 

 originally suspected. They are fasciate, not vittate ; and I have already 

 observed that Odonata never have fasciate legs, and might have added 

 that, so far as known to me, Ephemerina never have vittate legs. (See 

 Proa. Acad. Nat. Sci. Philad., Sept. 1862, p. 383.) 



It will have been noticed that Dr. Hagen calls the insect which I 

 sent him, and which was identical in every respect with the one figured 

 above, a larva and not a pupa. Authors generally tell us that in In- 

 secta the pupa is distinguishable from the larva by having rudimentary 

 wings, with the exception of course of those genera {Diaphcromera, 

 Phaphidophora, &c.) which have no wings whatever in the imago state. 

 Probably from the fact of there being no external wings in this insect, 

 as in all other known Pseudoneuropterous pupae, Dr. Hagen supposed 

 it to be in the larva state. The above, however, is only another ano- 

 maly in this most anomalous creature. I have a specimen in alcohol 

 from which the subimago has partly emerged, and which must neces- 

 sarily therefore be in the pupa state ; yet it has no external wings and 

 the subimaginal wings lie flatly under the interior surface of the cara- 

 pace; neither indeed had any of my other specimens any external wings. 

 It is, I think, a mistake to suppose that in lusecta the possession of rudi- 

 mentary wings is peculiar to the pupa as distinguished from the larva. 

 Many insects that I have bred, e. g. a Psyllade inhabiting a gall on 

 the hackhen-y (Celtis occidentalis), exhibit distinct rudimentary wings 

 before their final moult into the pupa state ; and I believe it is gene- 

 rally so with Pseudoneuroptera and Orthoptera, and probably in all 

 those Orders which have an active pupa. Unless, indeed, which is 

 contrary to all analogy, we choose to believe that an insect can moult 

 after assuming the pupa and before assuming the imago state, and so 

 consider it as a pupa as soon as the rudimentary wings begin to be 



