353] LARVAE OF THE TENTHREDINOIDEA—YUASA 35 



III. TAXONOMY 



Strictly speaking no classification of the Tenthredinoidea based upon 

 larval characters has hitherto been proposed. Attempts have been 

 restricted to the characterization of the different subdivisions included in 

 the superfamily. Among the earlier writers Le Peletier's (1823) work may 

 be mentioned. After a brief general account of the larvae, he gave a list 

 of eighteen divisions in which he grouped the species of the Tenthredinoidea 

 and stated whether the larvae of each division were known or unknown 

 and, if known, the number of the thoracic and abdominal legs present. 

 It is interesting to note that he mentioned a group of larvae the body of 

 which he characterized as "donkey-form" (aselliform) which he was 

 unable to place in any of his divisions. Dahlbom (1835) published careful 

 descriptions and a synopsis of larvae of sixty-three species. A synoptic 

 table for the larvae was compiled by Westwood (1840) from this work 

 and was published with additions. The characters used are the number of 

 abdominal legs and the feeding habits of the larvae. Norton (1867) 

 republished Westwood's table without additions. In the table given by 

 Cameron (1882) the larvae of more than ninety-five species are included. 

 The major groups are separated on the number of thoracic and abdominal 

 legs present. These subdivisions are segregated on biological characters 

 such as reflex bleeding, types of cocoons, and, finally, genera and species, 

 when known, are separated on the coloration, setae, food-plants, and feed- 

 ing habits. In 1895 Dyar, the most prolific and the only important Ameri- 

 can writer on the larvae of the Tenthredinoidea, published "A recognition 

 table for the known sawfly larvae of the North Atlantic States." The 

 larvae of one hundred and twenty-six species including forty-one not 

 specifically identified were considered in this synopsis. The characters 

 used are the number and location of abdominal legs, types of cocoon, 

 feeding habits, food-plants, and coloration. This last character was 

 employed extensively in separating different species. The next attempt 

 along this line was undertaken by Chester Young (1898), who was the 

 first to take into consideration the structural characters of the appendages 

 of the head. Unfortunately this work remains unpublished, but it is on 

 file as a baccalaureate thesis in the library of Cornell University. Konow 

 (1901) summarized the taxonomic information concerning the known 

 larvae of European and American species, four hundred and eighteen in 

 all, in the form of an analytical table. The presence or absence of abdom- 

 inal legs, number of antennal segments, and modifications and appendages 



