Ulmer's work in "Genera Insektorum" (1907) is the last version of his 

 classification, in which he still adheres to the traditions of the systematics 

 of caddis flies of his time: the order is divided into the suborders 

 Inaequipalpia and Aequipalpia. The artificiality of this division is even 

 more evident because it now contains not one but 4 families with case- 

 bearing larvae. 

 137 In an excellent work published in "Susswasserfauna Deutschlands" 



Ulmer (1909) described the order in all stages of development {egg, larvae, 

 pupae and imago) and gave his material in the natural order of the families, 

 from Rhyacophilidae to Sericostomatidae. He did not use, as before, the 

 artificial suborders proposed by Kolenati, but as in Siltala's classification, 

 Ulmer placed the Psychomyiidae next to the Phylopotamidae. 



Ulmer's classification (1907—1909) was widely accepted and was not 

 changed for a long time. Martynov (1924:18—20) following Wallengren, 

 proposed to consider the Beraeidae Wal. as an independent family; he 

 added 2 new families: Stenopsychidae (separated from the Philopotamidae) 

 and Arctopsychidae (separated from the Hydropsychidae). In his analysis 

 of the Stenopsychidae, Martynov also briefly described its characteristic 

 larva (Martynov, 1926b:291-292, Figures 9-14).* 



The larva of Arctopsyche ladogensis was also described 

 (Lepneva, 1928:28—32, Figures 1 — 11); the characters of this larva, which 

 show that it is more primitive than the larvae of Hy d r o p s y c h e, are 

 a strong argument in favor of the independence of the family. 



Martynov (1924:18— 20) included the results of studies Of the larvae of 

 Trichoptera during the last 40 years. He replaced the old division into 

 suborders by a new division. He stated (ibid., 18—19): "Kolenati (1848) 

 divided all Trichoptera into two groups (this division was modified in 1859): 

 Inaequipalpia, with an unequal number of segments of the maxillary palps 

 of males and females; and Aequipalpia, with 5 -segmented niaxillary palps in 

 both males and females. This division is artificial and does not agree with 

 the conclusions of Klapalek and Siltala which were based on the study of the 

 larva. Their study showed that the Trichoptera are already divided into 

 2 groups; 1 group has campodeiform larvae: (Rhyacophilidae, Hydroptilidae, 

 Philopotamidae, Polycentropodidae, Psychomyiidae, Hydropsychiidae), the 

 other has eruciform or suberuciform larvae (all other families). Great 

 differences in the biology of these 2 groups of larvae were found. Ulmer 

 also stated that the Trichoptera form 2 complexes of forms according to 

 the larvae." 



"The study of the wing venation, mouthparts and genitalia led to the 

 conclusion that the adult Trichoptera also form 2 suborders as clearly as 

 the larvae. The differences between these suborders are evident in the 

 various systems of organs; all this shows that despite the sometimes close 

 resemblance between the species of both suborders (which indicates parallel 

 evolution), the differences between them are marked and ancient. As the last 

 palpal segment is annulated in most groups of the first suborder,** I named 

 this suborder Annulipalpia and the second suborder Integripalpia, because 

 the last palpal segment is undivided. The Integripalpia show a close 

 relationship to Lepidoptera. " 



This larva was described in detail by Lepneva (1948a). 

 See p. 123. 



129 



