1917.1 43 



Collembola), rediiction of tracheae and the number of abdominal segments 

 (Collembolaj, and the occurrence of specialised organs like the caudal spring 

 and collophore of Collembola. The Apterygota seem too specialised to be in 

 the dirert line of descent of the winged forms (Pterygota). 



A diffei'ent concepticjn of a primitive genei-alised insect has been engen- 

 dered by the findings of Handlirscl) from the Palteozoic strata. As an 

 example, Stenodiitya is a large insect with well differentiated head, thorax, and 

 abdomen, with the full number (11) of abdominal segments, and with caudal 

 cerci. It has two pairs of large, long, and narrow wings, the venation of which 

 closely resembles that of the hypothetical wing suggested by Comstor]^ and Needham 

 sotne years before the pnbliration of Hatidli rsch's icorl:. Its prothorax also has a 

 pair of wing-like expansions of considerable size [c/., the nymphal stages of 

 certain Termites], and each abdominal segment has a pair of smaller lateral 

 expansions. In its early stages at any I'ate, Stenodictya had appendages on the 

 abdominal segments, but these may not have been pi'imitive, but secondarily 

 acqxiired gill-filaments like those of a Sialis larva, for Stenodictya was possibly 

 aquatic in its early life. 



If the fossil group Palaeodictyoptera, here exemplified by Stenodictya, 

 repi-esent the ancestral stock of I'terygote insects : what then were the 

 ancestors of Palaeodictyoptera ? From the available data may be constructed 

 a hypothetical ancestral insect, wingless with nervovis system less compacted 

 than in living forms, with paii-ed appendages on both thoracic and abdominal 

 segments (abdominal appendages are widely developed in the embryos 

 of living insects), and with a sei-ies of spiracles (structures which appear very 

 early in the embrj'onic life of living forms). The head is not discussed in 

 detail, but is considered as probably composed of six segments. Neither is it 

 possible to decide whether the individual appendages were biramous [two- 

 branched, the common Crustacean type] or not ; actual evidence only points to 

 a biramous condition in the case of the 1st and 2nd maxillae. 



The generalised ancestral insect being now visualised, where is the actual 

 creature most closely resembling it to be found ? Arachnoids are passed over 

 as too specialised. One turns to consider Trilobita, a group already highly 

 differentiated in the oldest knoAvn f ossilif erous rocks, and reaching their climax in 

 the Ordovician, the lowest great system ]>ut one; creatures which " dominated 

 the life of the oceans," occupying the ecological position now h<'ld by fishes. 

 With the increase of fishes in the Silurian epoch, the decline of trilobites began, 

 and they " became extinct with the passing of Palaeozoic time." More than 

 2(XX) spf'cies liave been described, doubtless l)ut a fraction of the whole 



It is sufficiently remarkable that it should have been possible to study in 

 detail the eyes and appendages, even the development, of creatiu-es which 

 ceased to live such long ages ago : but so it is, in the case of certain kinds. 

 The trilobite head has at least five pairs of appendages, and in Triarthrus it has 

 a pair of long filamentous antennae, very insect-like. Three kinds of eyes have 

 been found in trilobites, ocelli, aggregates of simple eyes, and true compound 

 eyes— corresponding roughly to the kinds in living insects. One trilobite genus 



D 2 



