438 M. N. Cliolodkowsky on the 



rest atrophy and finally disappear without leaving a trace 

 behind. Graber studied Oastropacha quercifolia^ in which 

 moth we are told that the abdominal appendages do not appear 

 until a relatively very late period (when the four cephalic 

 segments have become fused together), and then only on those 

 segments on which they are also present in the caterpillar ; 

 so that the series of embryonic abdominal appendages is not 

 a continuous one. From this Graber concludes, erroneously 

 referring to Tichomirow (the Russian text of Tichomirow's 

 paper was clearly unintelligible to him), that the pro-legs of 

 caterpillars are not homologous with the thoracic legs, and 

 represent secondary formations. At the same time he is 

 nevertheless prepared to allow that if a continuous series of 

 abdominal appendages is actually present in the embryo they 

 are liomologous with the thoracic legs. In my opinion there 

 is no question that the latter is actually the case ; for, in the 

 first place, the accuracy of Kowalevsky's assertion is hardly 

 open to doubt, since it is highly improbable that so cautious 

 and delicate an investigator as Kowalevsky, whose observa- 

 tions have almost without exception been confirmed by all 

 students, could go wrong in so simple a question ; and, 

 secondly, Tichomirow's statements also are very ample and 

 definite. It is true that Graber says that his drawings are 

 " indistinct ; " but only the appendages of the eleventh abdo- 

 minal segment are indistinct, or, strictly speaking, not shown 

 at all, in Tichomirow^s fig. 26, while the rest of the abdo- 

 minal appendages are indeed faintly outlined but perfectly 

 distinct. Particularly full and exact, however, is the descrip- 

 tion of the abdominal appendages given in the text (pp. 41- 

 42). To this I can further add that, on the basis of my own 

 investigations upon the germinal streak of Gastropacha pini, 

 I can entirely confirm Tichomirow's account, since in this 

 moth also at a very early stage in its development a con- 

 tinuous series of very small but yet distinct abdominal appen- 

 dages is observable, and the figure given by Tichomirow for 

 Bomhyx mori (fig. 26) also applies in every detail to Gastro- 

 pacha pini. 



But if, with reference to the external development of the 

 Lepidopterous embryo, we were absolutely unacquainted with 

 anything but the published observations of Graber upon 

 Gastropacha quercifolia , it would, I believe, nevertheless not 

 follow therefrom that the pro-legs of caterpillars are " secon- 

 dary " structures ; for whereon should sucli a conclusion be 

 founded — upon the late appearance of tiie abdominal appen- 

 dages or upon the fact that the series thereof is not an 

 unbroken one ? But late appearance in itself cannot be 



