34 CAUDAL GILLS OF ZYOOPTERID LARVAE, 



The scope of the present paper is sufficiently large to make it 

 advisable that it should be divided into four parts. Part i. deals 

 with the General Morphology of the Gills, and includes also a 

 Historical Summary of the work of previous authors, a list of the 

 material studied, and a short account of the biological methods 

 used. Part ii. deals only with the Morphology of the Separate 

 Gill-Types. Part iii. deals \\ith the Ontogeny, and Part iv. with 

 the Phylogeny of the Caudal Gills. It was originally intended 

 to publish the paper as a complete whole. However, certain 

 problems connected with Part iv. have made it imperative that 

 the Ontogeny should be studied in much greater detail than was 

 oi'iginally intended; in fact, it will be necessary to study by 

 sections all instars of the growing larva from the time of hatching 

 up to the attainment of the complete gill-form. To include this 

 would mean a delay of at least six months more. As Parts i. 

 and ii. include the principal results of more than eighteen months' 

 work, it seems best to publish these without further dela)% leaving 

 Parts iii. and iv. to appear together later on. 



In the Bibliography, placed at the end of Part ii., there will 

 be found all the publications known to me which deal with the 

 problems of respiration in Zygopterid larvje. Most of these have 

 little bearing upon the actual problem of the morphology of the 

 caudal gills, but it seemed advisable to offer as complete a list as 

 possible. References to the Bibliography are given in brackets 

 in heavy type. 



Part i. — General Morphology of the Caudal Gills. 

 Historical Summary. 



The first author to give a definite name to the caudal apjjend- 

 ages of Zygopterid larvje was Reaumur (26), who called t\\Qm Jins 

 ("nageoires"), but did not commit himself to any opinion of 

 their possible functions. A few years later, Roesel von Rosen- 

 liof(30) spoke of them as ^^rudder-feathers'' ("Ruder-Federn ') — 

 again, without making any suggestion as to their function. Both 

 these authors were evidently sioeaking of the common type of 

 gill found in Lestidcv and most Agrionidce, which I shall desig- 

 nate in this paper as the lamellar type. The first suggestion 



