1864.] 201 



As to Hemijttera, its mandibulate mouth excludes that Order. There remains 

 therefore only the order Orthoptera. of which there are only three families 

 witli aquatic larvse — Odonata, Perlina, and Ephemerina. The larvae of Odo- 

 nata always have the labium transformed into a well-known species of mask, 

 which is not found here. It cannot therefore belong to Odonata. Perlina 

 have only two caudal setas, instead of three. There remains Ephemerina, and 

 I believe that it belongs there, in spite of the antennae being, as you observe, 

 eight-jointed, and the absence of branchiae. Possibly, however, there are bran- 

 chife to the four basal segments of the abdomen, for I see something under the 

 carapace, but do not choose to ruin a unique specimen by a more violent exam- 

 ination. 



On inspecting the beautiful Ephemerina previously received from you, my 

 eyes accidentally fell upon Bcetisca obesa. Its robust form strikingly resembles 

 that of this larva, when viewed at a distance. After a minute examination I 

 believe that I am sure that this larva belongs to the genus Bcetisca, and proba- 

 bly to Bcetisca obesa; that is to say so far as one can be sure without actually 

 breeding the imago. The head and the oviparous lamina are alike, and the 

 carapace is represented in the imago, and even the groove on the dorsum 

 of the abdomen that fits into the tip of the carapace.* But there is a sort of 

 enigma here : for according to physiological and anatomical laws, we cannot 

 understand how the pro- meso- and meta-thorax can be all soldered together 

 in the larva. On the whole, it is about the most extraordinary larva known to 

 science. 



Those who are aware of the practical difficulty of correlating an in- 

 sect, known only in the larva or pupa state, with its imago, will appre- 

 ciate the successful acumen of the above analysis. One additional fea- 

 ture, by which the imago strikingly recals the larva and pupa, is not 

 referred to by Dr. Hagen. In the characters of the genus Bsetisca, I 

 noticed that '' the fifth abdominal joint is twice as long as any of the 



roptera to Orthoptera. See Monogr. Calopt. p. 1, note, and Monogr. Gomphin, 

 p. 1, note: also LeCoute's Introd. Class. Coleopt. p. viii, note. It is not quite 

 true, as suggested in the last passage by Baron Osten Sacken, that Pseudoueu- 

 roptera, as contradistinguished from Orthoptera, are "essentially aeriaZ, passing 

 the greater portion of the time on the wing." In Odonata, indeed, this is the 

 case, but Perlina and Psocina, and especially Psocina, to say nothing of Ter- 

 mitina, pass the greater portion of their time on trees, like the Orthopterous 

 Catydids and tree-crickets. 



* In the imago, as is usual in Ephemerina, the pro- and meso-thorax are sepa- 

 rated by a free suture, and the meso- and meta-thorax by a connate suture, but 

 the meso-scutellum extends over the abdomen to the tip of the first, or what 

 some would call the scrond abdominal joint, thus simulating the carapace of 

 the larva. The transverse, medial, sinuate carina on the 5th abdominal dorsal 

 is remarkably distinct and strongly recals that found in the pupa, though 

 it is not nearlv so much elevated. 



