URBICOLIDES (hESPERIIDES) . 87 



gradually develop under the epidermis during the growth of the larva 

 in its last larval instars, and are mature, so to speak, when the larva 

 is quite ready for pupation. They are placed on the underside of the 

 7th and 8th abdominal segments, and Riihl states that the excrescence 

 in Urbicola comma is " somewhat leaf -shaped, thick and viscous," that 

 " on the application of heat it melts like wax and therewith loses its 

 white colour entirely." It appears that these are in reality pockets of 

 some waxy material, which the larva scatters when spinning its 

 cocoon for pupation. It probably helps to make the cocoons 

 watertight, a very necessary provision, considering the slenderness and 

 usual position of the structures made by these caterpillars. 



The Urbicolid pupae vary much, and the main points of difference 

 between the various groups are most marked, e.g., Hesperiines, as 

 represented by the thick, stumpy, rounded pupae of Hesperia malvae 

 and Nisoniades tages ; the Urbicolines by the round-headed, long- 

 tongued pupae of Urbicola comma and Augiades sylvanus ; the Thyme- 

 licines by the sharp-nosed pupa of Adopaea Jiaca (thaumas), and the 

 Cyclopidid by the slender, beaked pupa of Cyclopid.es palaemon. The 

 mode of pupation in the Urbicolids is interesting. There is, first of 

 all, the making of a loose silken puparium, in which pupation takes 

 place. The Hesperiines {Nisoniades tages, etc.) appear not to have a 

 cincture or girth, although they cling most tenaciously to the inside of 

 the silken cocoon by means of the cremastral hooks, and it would 

 appear that the Urbicolines [Urbicola comma, etc.) also have none, 

 whilst the Thymelicids and Cyclopidids suspend themselves by an anal 

 pad and fine body cincture in the puparia they construct. Among 

 other butterflies w T hose larvae spin a cocoon are the Parnassiids, a 

 group of butterflies far removed from the skippers. 



Of the imaginal characters — to some of which reference has already 

 been made — the secondary sexual characters of the $ , the antennae, 

 the structure of the legs, and the neuration, are all characteristic. The 

 Megathymidae (Megathymus and JEgiale) are sometimes included in 

 the superfamily, as also is the Australian group, Eu&chemonidae (Eusche- 

 mon), in which the imago is furnished with a frenulum — a character 

 supposed to be absolutely distinctive of Heterocera. Their close 

 alliance with the Urbicolids is, however, very problematical. Of the 

 secondary sexual characters — (1) The costal fold of the $ as seen in 

 Hesperia malvae and Nisoniades tages, etc. (2) The discal " stigma" or 

 " brand " of the g asexhibited in Adopaea lineola, Urbicola comma, etc. (3) 

 The tuft of $ hairs as seen in Isoteinon atkinsoni, etc. — it is to be 

 observed that the costal fold is characteristic of the Hesperiids, the 

 discal streak or stigma to the Urbicolids (sens, restr.), etc. It may also 

 be here noted that in the Old World there are few species of Hesperiines 

 (i.e., with the costal fold on forewings) compared with the much greater 

 number of Urbicolines (i.e., with a discal stigma), whilst, in the New 

 World, the numbers of the two groups seem to be much more equal. 

 There appear to be species in both groups with other $ characters — 

 tufts and patches of modified androconial scales on underside of fore- 

 wing, or on upper- and underside of hindwings, or on legs, that is, 

 these are not peculiar to either group. Certain of the Urbicolid wing- 

 plumules, i.e., the special androconial scales found in these specialised 

 wing areas of the males, have been described and figured by Aurivillius 



