43 
neuration came under my notice the other day for the first time. It 
was a Cocytia which has been proved to belong to the Exotic family 
Hypsidae. The external appearance of the moth is strongly in 
favour of the Sphmgidae, but the positions of vein 5 of the forewing 
and vein 8 of hindwing clearly show that the insect cannot belong to 
that group on a neuration base. Neuration here has been proved to 
be correct, by the discovery of the larva, which is very like a Lymantriid 
larva and which establishes the position assigned to it. Although this is 
a remarkable instance there might be cited numbers of others less 
striking, and they suggest very strongly that we are on a sound basis. 
As has already been set forth by Tutt, families cannot be put as linear 
descendants one of another, and, in fact, Chapman’s work with the ova 
is proof positive of the impossibility of such an idea. Once this is 
clearly understood there is no doubt that many of the difficulties of 
neuration, difficulties brought about by attempting a direct lineal 
arrangement, will disappear. In the seven years that have elapsed 
since Meyrick first read his paper before the Entomological Society of 
London, our views have been immensely moderated, and many now 
obvious errors have been rectified. In the Entomologist's Record for 
1892 doubtless many of you will recall the storm aroused by Meyrick’s 
paper on the neuration. Tutt was particularly severe upon it, but must 
be credited with having said “ I quite agree that Mr. Meyrick’s facts 
will some day prove useful.” Many of the facts have already fulfilled 
that purpose, but the number of our entomologists who take any 
account of them is exceedingly small. Having endeavoured to uphold 
one particular character of the imago above others, I will try to show 
that the most valuable characters of the perfect insect are relatively of 
more account than characters of the larva. It is frequently evident 
that certain structures are quite useless to the insect and that they 
can easily do away with them. The larva of Philampelm satellitia 
affords a striking illustration. This North American Sphingid, which is 
a pest at times to the vine growers, has a larva which Avhen newly 
hatched has an immensely long caudal horn. After the first change 
of skin the horn is very much shorter, after the second, shorter still, 
and by the time the third is reached the horn has ceased to exist. We 
have here strong reason to suppose that in course of time the larva 
will have dispensed with the horn in all periods of its existence. 
It will be remembered that the caudal horn on the larva’s 12th segment 
is one of the strong family characters of the Sphinyidae. Why in this 
particular species the caudal horn should be taking the downward 
path is not known, or has perhaps not been thought out. Take again 
Acronycta ( Triaena ) tridens and Acronycta ( Triaena ) psi. With such 
closely related species it is reasonable to suppose that the larva? were 
structurally alike originally ; they are, however, now quite dissimilar, 
but the imagines of the two species (which must have been one 
originally) are so alike that few, if any, can always separate them. 
This from the facts to hand shows that the bifurcating from the 
common stock was most marked in the larva as it is there we have 
most divergence. If then these structures are plastic they are not so 
valuable as such characters in the perfect insect which are much less 
so. But there are other structural characters in the larva of which 
the tubercles are not the least important. 
These tubercles offer an immense amount of variation according 
