172 THE entomologist's record. 



)ii>iri)iata, and Eiiiatiir«ia atdinarla, large and yellow, very like our Cuxton 

 chalk-hill form, had evidently reached second hroods ; M. tristata, 

 too, could he disturhed from the hushes fairly high up, whilst Asjiilatcx 

 l/ilraria was frecjuent. Miaiia sti iiiilis was now and again disturhed 

 from a llower-head, having the hahit that one often ohserves with M. 

 fiinoinila on the Kent coast and elsewhere, of hasking in the afternoon 

 sun. Lithosia IntarcUa was hardly so golden-yellow as usual, hut still 

 had not reached typical L. jii/i/nutcola, as it has at Bourg d'Oisans, on 

 the other side of this mass of mountains. All the Setinas seen were 

 possihly S. aiin'ta, strongly streaked as usual. The Tineids and Tor- 

 tricids were positively ahsent, except for Hexersl Xantliosrtia ::(i('<;ana, an 

 odd Tortriv nilraiia, a few < 'atnjitiia sp. (/), and VicJirnrliainiiJut aliiina (/) , 

 w^hilst a single < iclcchia tacniolcUa appears to complete the hag. 



On the relationship of the Lepidopterous pupa to its larva. 



J5y A. BACOT. 

 The following notes have heen suggested hy some questions which 

 Mr. Tutt has submitted to me concerning the position of the hair-tufts 

 in certain Lymantriid (Liparid) pup^e and their relation to the hair- 

 tufts of the larva. The consideration of these hairs and the facts of 

 their position and strixcture raise some difficult questions with regard 

 to the theories hearing on the relation that the pupa bears to its larva. 

 T quite agree that the pupa must be considered as a separate and 

 independent structure apart from the larva or imago. Of course, 

 except as regards special, and prol)ably more recent, developments, the 

 pupa is nuich more ancestral than either the larva or imago. 



According to Weismann's theory of the mechanism which is present 

 in the germ-plasm of a lepidopterous insect (the only theory with 

 which I am sufhciently well acquainted to make use) we ought to have 

 four more or less distinct sets of determinants present, A — for the egg, 

 B — for the larva, C — for the pupa, D — for the imago. Theoretically 

 A and C should contain the oldest or most primitive determinants. 

 [Of course very old ancestral determinants such as produce the develop- 

 ment of sexual organs in the pupa, would be mixed with newer ones 

 (later developments), such as these accountable for the anal armature 

 (crunaster). These last-mentioned might possibly be far more recent 

 than those producing larval or imaginal characterr, while the former 

 would be necessarily much older than any imaginal characters, and 

 probably than many larval ones.] Detciminants A and D, for the 

 purposes of the present argument, may be dismissed. According to 

 the theory, netting aside scars and similar n:ere moulded or accidental 

 characters, where a character is ccmmon to larva and pupa (I am 

 speaking here cf characters of the pupal-case only, not of its contents, 

 as similarities of colour and pattern are due to the pupal contents), it 

 should be either an entirely independent development in both, or else 

 it naxst be considered older, phylcgeneticalJy, in the pupa than in the 

 larva. Hence it follows that, if the tubercles and hairs on the larva 

 undergo alteration cf structure or position, it by no means follows that 

 the pupal ones would l)e altered also. We ought, in fact, to find the 

 ancestral arrangement present in the pupa, in spite of an alteration! 

 being present in the larva. 



In a caddis-Hy larva the metathorax bears four dorsal plates, set im 

 trapezoidal form (there are also traces of trapezoidal groups en head,, 



