304 



PSYCHE. 



condition of these organs and what the 

 specialized conditions (for because de- 

 scent is not linear but branching, it is 

 illogical to speak of the most specialized 

 condition ; it is wholly possible for sev- 

 eral equally " most specialized " condi- 

 tions to obtain, each the apex of its own 

 kind of specialization [line of descent] ) . 

 Then are to be determined the lines or 

 tendencies of specialization, and the 

 intermediate conditions are to be ar- 

 ranged along these lines. After the 

 provisional determination of the homol- 

 ogies and phylogeny of the mouthparts 

 by the study of their comparative 

 anatomy, the development or ontog- 

 eny of the mouthparts of various flies 

 is to be studied, also with the aim of 

 determining the generalized mouthparts 

 condition and the paths of specializa- 

 tion. The results of the two methods 

 of study should be mutually confirma- 

 tory, if a correct interpretation of each 

 is reached. 



Of course the two methods of study 

 may, and often are, advantageously car- 

 ried on more or less nearly simultane- 

 ously, the revelations of one phase 

 helping materially to the quicker under- 

 standing of the phenomena of the other. 

 Or the ontogenetic study may precede 

 the comparative anatomical. Unfortu- 

 nately, in practice usually but one phase 

 of the study is prosecuted by a single 

 investigator, limitations of time, or ma- 

 terial, or of the capacity of the student 

 restraining him from the full double- 

 phased undertaking. 



Having begun the study of the dip- 



terous mouthparts some time ago, I 

 have progressed sufficiently to learn {a) 

 that the comparative anatomy of the 

 mouthparts is not an especially difficult 

 study, but that it alone may not certainly 

 determine the homologies of the dip- 

 terous mouth, and {h) that the onto- 

 genetic study of the mouthparts of 

 Diptera is an especially difficult study. 

 The Diptera exhibit " complete meta- 

 morphosis " in their life history. So 

 thoroughgoing is this metamorphosis, 

 as proved by the studies of Weissman, 

 d'Herculais, Viallanes, Kowalevsky, Van 

 Rees et a I, on the post embrj'onal devel- 

 opment of Musca, that the begin- 

 nings of almost all the imaginal organs 

 are to be looked for in the late larval 

 stages of life. The extraordinary his- 

 tolysis which is suffered by most of the 

 larval tissues and organs is so far-reach- 

 ing that most of the imaginal organs 

 develop from small groups of undiffer- 

 entiated cells, the imaginal buds or 

 histoblasts, which do not begin active 

 development until the fly has reached 

 and spent some time in, its larval stage. 

 This is conspicuously true of the integu- 

 ment and the appendages of the head 

 and thorax ; the wings, legs, and the 

 imaginal mouthparts arise from histo- 

 blasts whose development from groups 

 of invaginated hypodermal cells into 

 imaginal organs begins only rather late 

 in larval life. The difficulties of the 

 study of the development of the imagi- 

 nal appendages, and particularly of the 

 mouthparts, is difficult. The develop- 

 ment, beginning late, proceeds with 



