8^8 RADIAL AKD ZYGOtTEHlb SECTORS, kc, 



crossing M.^ again so as to lie between Mj and M2(Text-fig.2,6). 

 At the latest stage (Text-fig. 2,c), Rs is thrown across both Mj 

 and M2, so as to lie between M^ and M3. 



Dr. Ris has photographed (Textfig.2,rf) the triangular wing- 

 bud in a very early stage of Lihellula (this would appear tu be 

 one instar earlier than Needham's earliest stage, since the wing 

 is only 08 mm. long, and definitely more triangular in shape 

 than Needham's wing of Gomphus, which was 1 mm. long) This 

 shows Rs already crossing both Mj and Mg! I find the same 

 result in the earliest stages of Hemicordulia. But we must 

 remember that the Libelhdidfe are far more specialised than 

 Gomphus. The latter (if Needham's observations are correct, 

 and I see no reason to doubt them) has preserved the ontogenetic 

 stages in detail; the former reaches the final condition of Rs 

 while the wing is yet very small. 



We see, then, that the ontogenetic evidence supports the 

 structural evidence, but with greater force. IJoiv can a trachea 

 which, ill the earliest examinable larval stage of Anisoptera, is a 

 branch of R tiot eiien crossing M at all, and ivhich only reaches its 

 fitial position (in Gomphus) by two successive changes, be homolo- 

 gous with the branch of M ivhich is present in a fixed position 

 (between M^ and M3) in Zygopt- ra fiom the earliest examinable 

 stage! One must either accept the proof as definitely against the 

 homology, or deny one's belief altogether in the Biogenetic Law. 



We must conclude then, both on the structural and on the 

 ontogenetic evidence, that trachea lis of Atrisnptera is not homo- 

 logous ivith trachea Ms of Zygoptera. 



3. The Evidence from Bridges and Oblique Veins. 

 , We have now to approach very cautiously the crux of the 

 whole problem, viz., whether we have been correct in maintain- 

 ing that the imaginal veins Ms and Rs are wholly laid down 

 upon the preceding tracheae Ms and Rs respectively, or whether 

 they may not have, in special cases, a more composite origin. 



Thus we have to deal with a question that is even more funda- 

 mental than the point immediately at issue, viz., the interaction 

 of the two great forces whose interplay has moulded the Odonate 

 wing from the very start. These are, the force of tracheational 



