EMBRYOLOGY OF CHRYSOPAj ETC. 141 



but Newport's observations on Jiilus indicate that it is deve- 

 loped in an entirely different mode from the Hexopoda or 

 Arachnida, a remarkable feature being the persistence of the 

 larva in its inner (?) embryonal membrane {faltenhlatt of 

 Weissmann) for many days after it is hatched. 



There are, however, two modes of development in the 

 Hexapoda, depending on the position of the primitive band 

 in relation to the yolk. The Hymenoptera, Diptera, and 

 certain Coleoptera (Curculionidse), and Donacia, and the 

 Phryganeidse and Podurse (Isotoma) are ectohlasts} while 

 Telephorus and the Hemiptera and certain Neuroptera 

 (Libellulidae and Hemerobidae) are enclohlasts, to use Dr. 

 Dohi'n's terms. On inquiring how far these two modes 

 correspond to the degree of development of the insect on 

 leaving the e^y^, and the degree of metamorphosis of the 

 insect before becoming adult, it seems that the endoblasts 

 occur in those ametabolous insects (Hemiptera and Neuro- 

 ptera) with flattened, leptiform larvae, and also in those 

 Coleoptera with similar larvae, as distinguished from the 

 weevils, which have cruciform larvae, i.e. resembling the 

 maggots of Diptera and Hymenoptera. The two modes of 

 development, then, do not fully accord with the two different 

 degrees of metamorphosis of insects, but more probably 

 depend simply on the form of the larva when hatched. Now, 

 there are two forms of insect larvae which are pretty con- 

 stant. One we may call leptiform, from its general resem- 

 blance to the larvae of the mites (Leptus). The larvae of all 

 the Neuroptera, except those of the Phryganeidae and Panor- 

 pidae (which are cylindrical and resemble caterpillars), are 

 more or less leptiform, i.e. have a flattened or oval body, 

 with large thoracic legs. Such are the larvae of the Ortho- 

 ptera and Hemiptera, and the Coleoptera (except the Curcu- 

 lionidae ; possibly the Cerambycidae and Buprestidae, Avhich 

 approach the maggot-like form of the larvae of weevils). On 

 the other hand, taking the caterpillar or bee larva with their 

 cylindrical fleshy bodies, in most respects typical of the larval 

 forms of the Hymenoptera, Lepidoptera and Diptera, as the 

 type of the eruciform larva, we find that those insects with 

 such larvae are ectoblasts. (The Podurae which, as in Iso- 

 toma, are ectoblasts, and are certainly leptiform when 

 hatched, form an apparent exception.) Thus, the two modes 

 of development (ectoblastic or endoblastic) perhaps simply 



' I omit any reference to the Lepidoptera, which Dr. Dohrn regards as 

 endoblasts, but which I am inclined, from some eggs (probably of an Arctian) 

 I have studied to regard as developing like the Hymenoptera and Diptera. 



VOL. XII. — NEW SEE. K 



