SECT. 2] 



AND WEIGHT 



509 



the nature of the Hmiting factor or master process, in view of the 

 fact that Q^^Q itself was not a satisfactory standard. When a physical 

 process of diffusion (Nernst) could have a coefficient practically 

 identical with a chemical process (saponification of ethyl butyrate 

 by baryta, Trautz & Volkmann), both being about 1-3, it was difficult 

 to say what might be producing a given Q,io • But more serious was 

 the fact which soon became apparent, namely, that many cases of 



embryonic development could equally well be said to be simply 

 linear functions of the temperature. Krogh was the first to find this. 

 In considering the figures which had been obtained by Krogh & 

 Johansen and by Dannevig on the rate of development of fish eggs 

 and the hatching times at diflferent temperatures, he found that the 

 relation with velocity was a straight line and not an exponential 

 curve, and that the relation with time was a curve which was 

 not exponential but was a hyperbola. This is seen in Figs. 78 and 



