MEMOIliS OF THE NATIONAL ACADEMY OF SCIENCES. 19 



liviiis' in trees or shrubs several or many feet above the gronTid are certainly exposed to a more 

 even temperature, as it is colder at ni,i;ht even in midsummer within a lew inches of the ground, 

 say about a foot, the usual lieij;ht to which grasses and herbs grow. The changes, therefore, by day 

 and night are greater at the surface of the ground than among the leaves and Inaiudies of a tree. 

 ;SI(n-eover, forests, not too dense for ins(H't life, with glades and paths to admit tiie sunlight and 

 heat, must necessarily have a more even temperature and be less cxijosed to cool winds, aud less- 

 subject to periods of drought than grassy fields. There is also a less free circulation of air among 

 grasses and herbs, which may be more or less matted and lodged alter heavy rains, than among the 

 separate and coarser leaves of trees, such as the different species of oak, which in North America, 

 at least north of Mexico, harbors a far greater nn.mber of species of insects (over .500) than any 

 other plant known. On the whole, forest trees support a far larger number of kinds of phytopha- 

 gous insects than grasses or herbs, and may this not be due to better air and a freer circulation, 

 to a more equable temperature, perhaps of a higher average, and thus lead insects to eat more? 

 May not the plump bodies of the larger silkworms, as the larval Attaci, the Ceratocampids, and 

 especially the Cochliopodida' (Limaeodes), be in some way due to their strictly arboreal envi'-on- 

 ment?' 



When the ancestors of the inesent groujjs became fairly established under these changed 

 conditions, beconung high feeders, and rarely wandering to low herbaceous plants, we should 

 have a condition of things akin to geographical isolation. The species would gradually tend to 

 become segregated. The females would more and more tend to dejiosit their eggs on the bark or 

 leaves of trees, gradually deserting annual herbs. 



For example, the females of the Attaci and their allies, as well as the Cochliopodida', may have 

 at first had larger wings and smaller bodies, or been more active during flight than their descendants. 

 Their present heavy, thick bodies and sluggish habits are evidently sec(mdary and adaptive, and 

 these features were induced perhaps by the habit of the females ovipositing directly ui)on leaving 

 their cocoon, and cocoon-spinning moths are perhaps as a rule more sluggish and heavy-bodied 

 than those which enter the earth to transform, as witness the Ceratocampid?e compared with 

 the cocoon-spinning silkworm {B. mori) and the Attaci. Spinning their cocoons among the 

 leaves at a jieriod in the earth's history when there was no alternation of winter and summer 

 and probably only times of drought, as in the dry season of the Tropics at the present day, the 

 females may have gradually formed the habit of depositing their eggs immediately after exclusion 

 and on the leaves of the trees forming their larval abode. The females thus scarcely used their 

 wings, while (as in Cdltosumia iiromeihcn) the males, with their larger wings, lighter bodies, 

 broadly pectinated antenna^, and consequently far keener sense of smell, could tly to a greater 

 or less distance in search of their mates.- The principal of segregation-' so well worked out by 

 Mr. Gulick, to which Mr. Romanes' theory of physiological selection is a closely allied factor, if 

 not covering the same ground, would soon be in operation, and the tendency to breed only among 

 themselves, rather than with the low feeders, would more and nn)re assert itself, until, as at present, 

 arboreal moths, as a rule almost, if not wholly, oviposit exclusively on the leaves or bark of trees. 



' The fat, overgrown slugworms (Limaeodes) may be compared to the overfed, high-bred pig, whieh eats- 

 voraciously, has little need of routing, and takes l)ut little exercise. Where, as among cave Vininials, there is a. 

 deticiency of food, we have a constant tendency to sliniuess, to an attenuation of the body. This is seen in the 

 blind cave arthropods, such as the blind crayfish, blind beetles, blind C:pci<lotipa, etc., compared with their allies 

 which live under normal conditions. (.See the author's memoir on the Cave Fauna of North America, etc., Mem. 

 Nat. Acad. Sciences, iv, 24, 1889.) 



= The secondary sexual characters so marked in Bombyces are perhaps the result of their peculiar arboreal 

 habits; so also the apterous tendency of Orgyia and a few other forms, especially the arboreal Psychidie (fEceticm 

 and ThiirUlopterijx), as well as Aniaoptcnjx aud Ilibcrnia. The larva- of the Xi/ssia feed on trees or low plants. It 

 may be i)uesti()neil ^vhether any wingless female I.ei)idoptera live on herbaceous jilants. Contrast with them the 

 grass-feeding species of Noctuiihe, a:j those of .Igroiin, Leucania, etc. 



'In fact nearly the whole group of insects is an example on a vast scale of the principles of segregation, 

 geographical isolation, and physiological si'lection. As soon as the ancesters of insects acquired wings their mUku 

 was changed. The air rather than the earth became iheir habitat; the accpiisition of wings introduced tlicm to a 

 new world of existence, and free from the atta<'ks of creeping enemies and other adverse conditions to which the 

 terrestrial Myriopods and Arachnids were subjected; the winged insects living a part of their lives, and the most 

 important part, above the surface of the soil, multiplied prodigiously, the iiuuiliir of species being estimated by 

 millions when we take into account the fossil as well as the living forms. 



