Ulmer, Siltala and Martynov published their work on the larvae of caddis 

 flies almost at the same time (1900—1902); Thienemann published his work 

 on caddis fly larvae in 1903. 



Martynov described the peritrophic membrane of caddis fly larvae and 

 their glands in his first publications; he rarely described the larvae of 

 Trichoptera in his later work, giving only some short descriptions of 

 larvae occasionally (Martynov, 1901; 1903; 1926a:24-28, 31-33, 35-56; 

 1926c:291— 292). He used the data which had accumulated to substantiate 

 his new division of the Trichoptera into suborders (Martynov, 1924a:18— 19). 



Ulmer studied the larvae of Trichoptera for 10 years;* later he also 

 published keys and surveys (Ulmer, 1912a; 1925a; 192 7); the 2 volumes of 

 his monograph contain numerous descriptions of larvae (Ulmer, 1955, 1957). 



On the basis of his material (about 80 species) and on that of Struck, Klapalek, 

 Morton and Siltala, Ulmer published a large survey of the caddis fly larvae 

 of Europe; this work contained a detailed morphological section, keys and 

 an exhaustive bibliography (Ulmer, 1903). The work is based on the division 

 of McLachlan's 7 families into different groups, replacing McLachlan's 

 sections with subfamilies which are based also on the morphology of the 

 larvae; the family Limnophilidae was divided into the subfamilies Limno- 

 philinae and Apataniinae; the family Sericostomatidae was divided into the 

 subfamilies** Sericostomatinae, Goerinae, Brachycentrinae and Lepidosto- 

 matinae; the family Leptoceridae was divided into the subfamilies Beraeinae, 

 Molanninae, Odontocerinae and Leptocerinae; the family Hydropsychidae 

 133 was divided into the subfamilies Hydropsychinae, Philopotaminae, Polycentro- 

 podinaet and Ecnominae. Some of these subfamilies correspond to the 

 families of Wallengren; this detailed classification was also applied to the 

 larvae. 



It is no accident that the division into the artificial suborders Inaequi- 

 palpia and Aequipalpia does not appear in Ulmer's book (1903); he was an 

 expert on caddis fly larvae, and it is impossible to give a diagnosis of these 

 artificial groups for the larvae. In his key to families and subfamilies, 

 Ulmer (1903:30—33) divides the order according to the larvae into the 

 natural groups established by Klapalek, namely campodeiform and eruciform 

 larvae; he placed the group of suberuciform larvae between them; Ulmer 

 attempted to maintain, as far as possible, the system accepted in his time 

 for the adult caddis flies; therefore, he accepted the same sequence of 

 families as McLachlan, from Phryganeidae to Rhyacophilidae; he placed the 

 third of the above groups at the head of his system, then the second and 

 finally the first (Ulmer, 1903:30-31). The key to families based on the 

 morphology of the larvae reflects the phylogenetic relationship of the groups 

 better than keys of adults. Ulmer (l903:3l) introduced an important 

 correction into this key: he separated the family Leptoceridae with its 

 eruciform larva from the families with campodeiform larvae, and placed 

 it in its natural position in the group of families with eruciform larvae. 

 The artificial classification of the Leptoceridae in the Aequipalpia Kol. was 

 again stressed by Siltala. 



* Ulmer called this period "Meine Larvenperiode." 

 *" Ulmer retains the classification of McLachlan. 



t In this work, and in his later work (Ulmer, 1906, 1907) wrote Polycentropinae and Polycentropidae as he 

 established this subfamily and family; later, taxonomists, including Ulmer, changed these names into 

 Polycentropodinae and Polycentropodidae. We use these last names in the text and when we quote authors 

 who did not use the names. 



125 



