June 1899.1 CaSEV : On AMERICAN CoCCtNEl.l.lD/E. 71 A REVISION OF THE AMERICAN COCCINELLIDiE. Bv Thos. L. Casey. The object of the following pages is to give a short outline or sketch of every species occurring within the limits of the United States accessible to me at the present time, and also to invite attention to certain features in the taxonomy of the family which do not seem to have been hitherto brought to notice. In an appendix a list of African species is given, containing quite a number of novelties, and the descriptions of certain new species from other parts of the world are also appended. COCCINELLID^.. The separation of this family into two parts based upon mandibu-lar structure has never seemed entirely satisfactory to me ; first, because of the difficulty of observing the character, causing the classification of Chapuis to be unpractical, and, secondly, because Epilachna and related genera are merely pubescent halyziids, slightly modified by reason of perverted food habits and attendant environments. Many of the Harpalini of the Carabidte are known to be either Avholly or partially phytophagous, but no one has proposed to divide the Carabidte on these lines, and would scarcely do so even if a minute structural divergence in the mandibles existed, and it has never been demon-strated that the mandibular teeth serving as the basis of the Chapuisian classification are not found elsewhere in the family. The Epilachnini, in fact, resemble the Psylloborini in all external structures, including the long antennae, a character of more importance than has apparently been conceded. In view of these facts I have not employed the classi-fication of Chapuis in the following pages. The latter author appeared also to be constantly striving to reduce the generic groups hitherto proposed, but this cannot be done with pro-priety, and many more will be needed, both of genera and tribes, before the taxonomy of the family can be made entirely clear. This is well shown by some small species which we had held to belong to the genus Peutilia, until Weise recently proved that they were in no way re-lated, and separated them under the name Sinilia ; as a matter of fact they do not resemble Pentilia at all, and are much more closely allied
BioStor
