METHODS OF MEASURING RESPIRATORY EXCHANGE 21 



Respiration experiments on aquatic animals breathing dissolved 

 oxygen differ greatly in technique, though not in principle, from the 

 corresponding experiments on air-breathing forms. I shall make this 

 last difference the basis of my classification and describe : 



1 . The methods for studying respiratory exchange in air, and 



2. The corresponding methods to be used with water as the re- 

 spiratory medium. 



I do not propose to describe in technical detail the large number 

 of respiration apparatus constructed, 1 but I shall endeavour to charac- 

 terize the different types of apparatus and to indicate their proper 

 sphere of usefulness and their limitations. 



Within each type we have a number of instruments, composed of 

 essentially the same principal parts (e.g. air circulators, CO 2 absorbers, 

 etc.). Each part exists in a variety of forms which differ greatly in 

 their technical appearance, though they are intended for the same 

 purpose. I do not propose to describe all or nearly all such forms 

 but to select those which in my judgment are best adapted for their 

 purpose and to indicate as far as my experience goes the most 

 useful combinations. My descriptions are intended throughout as 

 guides for the selection of instruments but not for their technical 

 construction. 



. The methods for studying respiratory exchange in air fall natur- 

 ally into two groups : 



i. A. The methods by which the total gas exchange is determined 

 while the animal under experiment is enclosed in a suitable respiration 

 chamber but otherwise not interfered with, so that the conditions may 

 approach the normal life as nearly as possible. These methods can 

 be applied to air-breathing animals of all classes and all sizes. They 

 are especially suitable for experiments of rather long duration from 

 about an hour upwards. 



I. B. The methods by which the specific respiratory organs only 

 are connected with a respiration apparatus. These methods will 

 almost invariably interfere greatly with the freedom of movement of 

 the animal. They are suitable therefore for determinations involving 

 definite experimental conditions, e.g. narcosis, operations, etc. In 

 practice they can be applied only to vertebrates of a certain size and 

 over comparatively short periods only. A number of these methods 



1 Such descriptions must be sought in the original papers referred to and in Tigerstedt's 

 " Handb. der physiol. Methodik ", 



