128 MR. R. J. TILLYARD ON THE RECTAL BREATHING-APPARATUS 



complexitj' o£ the gill-formation made the investigation in almost every case 

 not only a difficult one, but very long and tedious. 



In order to avoid prolonging the investigation unduly, I have found it 

 necessary to postpone not only the physiological study of the respiration, as 

 carried on in the rectum, but also to confine my attention almost entirely to 

 mature or fairly mature nymphs. A complete phylogenetic study demands, 

 of course, a thorough investigation also into the successive ontogenetic 

 changes that take place at each ecdysis throughout the growth of each larva 

 studied. This would, however, be a matter not of months merely, but of 

 years. In the twelve months during which this work has been carried out, 

 I was only able to study the ontogenetic gill-development in two genera 

 {^iEschna and Aiiax) — fortunately, genera which throw much light on the 

 phylogenetic problem. The object of the paper may therefore be briefly 

 stated as an attempt at a complete morphological study of all obtainable 

 forms of rectal breathing-apparatus in Australian Anisopterid dragonfly 

 nymphs, with some phylogenetic deductions which appear to me to be fully 

 warranted by the facts therein established. 



Some of the facts made known for the first time in this paper do 

 undoubtedly throw light upon the very difficult physiological questions 

 involved. It seems to me, however, to be unsafe to draw physiological con- 

 clusions, as Sadones (25) has done, from purely morphological and histological 

 studies. I therefore prefer to hold over this aspect of the question until a 

 later date, when perhaps physiological experiments may have yielded more 

 conclusive evidence than can be to-day offered in support of any theory of 

 respiration in these insects. If at any point I have touched upon the 

 physiological aspect of the problem, it is because that aspect happens to 

 stand out very clearly in relation to the facts in hand, and not because they 

 offer anything like a complete or satisfactory solution of the problem. 



I should like to express my thanks to Professor W. A. Haswell, F.R.S., 

 under whom this research has been carried out, for much kindly encourage- 

 ment and sound advice ; while to Dr. F. Ris I am indebted for the 

 origination of the subject and also for valuable criticism during the progress 

 of the research. 



Historical Summary. 



In the appended Bibliography there will be found a complete list of all 

 the authors known to me who have mentioned the rectal breathing-apparatus 

 of Odonate larvse. These peculiar rectal structures seem to have been 

 noticed first by Swammerdam. (28), though the actual recognition of them as 

 a breathing-organ is probably due to Poupart (20). The respiratory function 

 of the rectum was also known to Reaumur (22), de Greer (9), Cuvier (3), 

 Duvernoy (6), Marcel de Serres (14), Lyonnet (13), Suckow (27), and 



