154 THE CANADIAN ENTOMOLOGIST. 



species do not resemble those of the latter family. The caterpillars of the 

 Cuspidates are singular in form, many of them being without anal pro- 

 legs, and others being bifid at that extremity. Very few American species 

 have this formation, but some of them are humped, and are otherwise of 

 uncommon shape, while others have the cuspidate habit of raising the 

 anal extremity in the air while feeding or moving about. The sub-family 

 includes the well known genera, Icthyiira, Datana, Notodonta, Ceriira 

 and others. 



Platvptericid.«. This division is included in the European family 

 Cuspidates. The moths are small, and all have falcate or hooked wings. 

 In England the caterpillars of these moths are not hairy, but are marked 

 with colored lines and spots, and some are without anal prolegs. 



Attaci. Comprises our largest and most beautiful moths, as well as 

 those which produce the best and most plentiful supply of silk. Both 

 moths and larvge are well known. 



Ceratocampid<«. The genera Dryocampa, Hyperchiria, Anisota and 

 others belong to this sub-family of large and handsome moths. Some of 

 the larvje make no cocoons, but bury themselves in the earth like the 

 Sphingidae, and there become pupae. 



Lachneides. Includes Gastropac/ia, Clisiocampa and Tolype, moths 

 of medium or small size, with hairy bodies. The larvge are smooth, with 

 few hairs, and spin shght cocoons of silken fibre. 



Hepialid^. a group of wood or root-boring moths, some of which 

 are large and robust, others small and delicate in appearance. 



A glance at the foregoing sketch will show the variety of insects we 

 include in this family, and also the widely diff'ering habits of the larvai. 

 And it is both curious and interesting to note the analogies which many 

 of the species bear to the other divisions of the Lepidoptera, and even to 

 the other orders of insects. In some cases the resemblance is so close 

 that it is doubtful whether the species is rightly placed, and we might be 

 justified in removing it to some other family of moths, perhaps a long way 

 from the Bombycidte. Thus in the first sub-family, the Lithosiina;, the 

 genus Cranibidia very much resembles the genus Crambus among the 

 rineidse. Euphanessa nmidica is very like a Geometer, not only in the 

 perfect state, but also as a larva, as stated by Mr. Saunders, who calls it 

 a ^' true looper." In a classification based on larval characteristics, this 

 moth would therefore be ranked with the Geometridse. Crocota ferru- 



