380 LIBBIE H. HYMAX. 



of the type of gradient which we have described for it. The 

 acetone powder experiments do not, in my opinion, answer these 

 objections. If the substance in such powders which absorbs 

 oxygen is essential to respiration then it is necessarily true that 

 types of tissues which have a high respiratory rate must contain 

 a higher percentage of the substance. It has been repeatedly 

 emphasized that direct proof of axial respiratory differences can 

 be obtained only through the comparison of pieces of like morpho- 

 logical constitution and through the elimination of various func- 

 tional factors which affect metabolic rate. It is to be hoped 

 that Shearer will repeat these experiments with more suitable 

 material. Such material is necessarily limited to the lower 

 invertebrates or to the very early embryonic stages of higher 

 forms. 



The sponge Grantia seemed to me to constitute very favorable 

 material for a further test of the reality of axial metabolic 

 di (Terences. It has the same morphological constitution through- 

 out except at osculum and base and is more or less definitely 

 polarized. The work w r as performed at the Marine Biological 

 Laboratory at Woods Hole during the summer of 1924. 



i. Method of Determining the Oxygen Consumption. For two 

 or three years I have been trying to devise an apparatus suitable 

 for the study of the oxygen consumption of small organisms by 

 Winkler's method. The method finally adopted owes its origin 

 to a device described by Osterhaut and Haas ('17). They first 

 suggested an apparatus separable into two pieces, one part to 

 contain the organisms and the other part for analysis. Their 

 apparatus is, how r ever, clumsy to manipulate and for large 

 animals or large amounts of material the method used by me for 

 many years of siphoning off the sample is very much simpler 

 and entirely satisfactory, as proved by checks. The device pro- 

 posed by Osterhaut and Haas to prevent exposure to air in adding 

 the reagents is in my opinion wholly unnecessary unless one is 

 dealing with water of very low oxygen content. In working 

 with small organisms, a very much smaller apparatus is required. 

 This naturally reduces the size of the sample of water available 

 for analysis. This difficulty is overcome by using a smaller 

 quantity of the reagents and a more dilute thiosulphate solution 

 for the final titration, as also suggested by Lund ('22). 



