.8,,. THE DEVELOPMENT OF INSECTS. 281 



been regarded as an instance of atavism, of retrogression towards 

 the form of an extremely remote progenitor. 



The primitive nature of the campodiform larva has, however, now 

 been called in question by Cholodkowsky. In the concluding section 

 (on the morphology and phylogeny of insects) of his recent paper (3) 

 he reviews the latest researches on insect embryology. He states 

 that Kowalevsky and Tichomirow have described appendages on the 

 abdominal segments of the embryos of moths {Smerinthus populi and 

 Bombyx mori). Graber also described abdominal appendages in the 

 embryo of Gastropacha pini, but considered them secondary structures 

 because of their late appearance. Now, Cholodkowsky believes that 

 these embryonic appendages disprove the derivation of the caterpillar 

 from the campodean larva, and establish the homology of the pro-legs 

 with the legs of the thorax. On the other hand, he does not consider 

 the vermiform larva to have given rise to the campodean, because the 

 most primitive winged insects (Orthoptera) have only six legs in their 

 young stage. The conclusions seem to be, that the two kinds of larva 

 have arisen independently of each other, that all insects are derived 

 from a remote many-footed vermiform ancestor (as forms with a 

 campodean larva have abdominal appendages in the emhryomc stage), 

 but that only some groups can be descended from a campodean 

 ancestor. The entire class of insects is more nearly related to the 

 Myriapods than to any other group of Arthropods. 



Many questions of great interest are discussed in this work of 

 •Cholodkowsky. He has observed appendages on all the eleven 

 abdominal segments of the embryo cockroach, and he concludes that 

 all these are homologous with the legs and jaws. The appendages 

 of the four hinder abdominal segments become the gonapophyses and 

 cerci of the adult insect, the cerci being the limbs of the hindermost 

 ■segment. It hence appears that these organs had an ambulatory 

 function in the remote worm-like ancestors of the class. 



One of the most startling of Cholodkowsky's conclusions is his 

 interpretation of the insect-head. He states that this consists of 

 ■certainly more than four, probably six, segments of which one only 

 is pre-oral, and that the antennae are the appendages of the first post- 

 oral segment. The pre-oral origin of the antennae of insects has 

 •seemed so firmly established during the last few years that this 

 statement will not be received without further confirmation. The 

 •contradictory statements already made by various observers on this 

 point are proof of the difficulty of determining whether the embryonic 

 antennae really arise before or behind the mouth, and Cholodkowsky's 

 ■own figures do not seem conclusive on the subject. However, he 

 states that there is doubtless an antennal mesodermal somite (clearly 

 figured), and that the rudiment of the antennal ganglion arises behind 

 that of the pre-oral ganglion with which it afterwards fuses. 



Should this post-oral origin of the antennae of insects be estab- 

 lished, they will have to be regarded as the homologues of one of the 



