Introduction 27 



bright colors, and arboreal habits seem to constitute an isolated group, 

 perhaps even higher in rank. 



The abdomen has always six or more segments in Adephaga, seven 

 in Cicindelidse cf and in Gyrinidae, eight in the genus Brachinus. The 

 tarsi are 5-jointed throughout, but in Hydroporus the front and middle 

 tarsi are apparently 4-jointed, the fourth joint being either actually 

 wanting or concealed by the deeply lobed third joint. Ocelli are wanting 

 in the adults, but very perfect in the larvae of Cicindelidae. In the Cara- 

 bidse a striking peculiarity is the development of tactile setae. These 

 are wanting in the genus Oodes and the aquatic Adephaga; also in the 

 subfamily Pseudomorphinae, which is also remarkable for its short legs 

 and rigid tarsi. Still another large group is characterized by fossorial 

 legs, by which it aids its underground operations, and by pedunculate 

 thorax. The position of Amphizoa is a matter of doubt; but I have 

 not attempted to make any changes in the place at present assigned 

 to it or other divisions. The sequence follows Americto precedents 

 because there is no other at present available. 



POLYPHAGA 



The number of families in this sub-order is so great that it wU be 

 most convenient to consider the divisions. I use the terms that have been 

 suggested by Lameere principallv, adding Mordelloidea for the remainder 

 of the old series Heteromera, after separating Tenebrionoidea. 



PALPICORNIA or HYDROPHILOIDEA 



The great length of the palpi, exceeding that of the antennae 

 in the most derivative forms, but far less developed in the primitive 

 sub-families, gives this series its name. In Dr. Leconte's system the 

 principal families included follow the Adephaga, and I have made no 

 alteration. The campodeaform larva of the Hych-ophilidae seems to sup- 

 port Dr. Leconte's view. The phylogeny has been carefully studied by 

 d'Orchymont;' he arranges the sub-families in the following order, viz: 

 Hydraeninae, Limnebiinse and Spercheinae (not American) as the more 

 primitive, and Helophorinae, Ei)imetopinae, Hydrochinae, Sphaeridiinae, 

 Hydrophilinae, as the more derivative; and agrees with Handlu'sch and 



1 Ann. -Soc. Ent. Fr. LXXXV, 1916, pp. 91-106; and 23.5-240. 



