714 THE RESPIRATION AND [pt. iii 



of consideration the early production of carbon dioxide by the shell. 

 His experiments differed from those of his predecessors in that he 

 made no attempt to incubate the eggs in the respiration apparatus, 

 but studied their carbon dioxide output for short periods removed 

 from the incubator, from the 6th day onwards. Ha\ing obtained a 

 final set of figures, he related the cubic centimetres of carbon dioxide 

 put out per 24 hours to the log. weight (as shown in Fig. 153), and 

 then, reading off on the ordinate the amounts of carbon dioxide 

 corresponding to the log. weights at definite days, obtained a graph 

 relating carbon dioxide production to age. This is naturally more 

 accurate than proceeding directly. The curve so obtained is shown 

 in Fig. 151, together with those of Bohr & Hasselbalch, of Atwood & 

 Weakley, and of Hanan. Taking the data as a whole, it is evident 

 that, for the most part, the values of the American workers are higher 

 than those of the Danish ones; this is probably a difference of breed. 

 The beginnings and the ends of the curves are interesting. All 

 the curves which go as far back as the ist day show the small output 

 of carbon dioxide from the shell, and all of them, except the second 

 series of Bohr & Hasselbalch, show the falHng off about the 

 17th day, which appears also in the sigmoid form of Murray's carbon 

 dioxide per log. weight cur\'e just given. For an explanation of this 

 phenomenon, see Brody (p. 431). Murray's curve begins slightly 

 below the Danish ones, then between the 7th and 17th days moves 

 up above them, suffering a temporary eclipse on the i8th day. 

 Murray did not offer any explanation for these small di\'ergences. 

 In a later paper he calculated the metabolic rate from these carbon 

 dioxide figures, and in Fig. 154 is shown the cubic centimetres of 

 carbon dioxide produced per gram embryo (wet weight) per day 

 plotted against the age from 5 to 19 days of incubation. The Minot 

 growth-rate curve is included to illustrate the diphasic character of 

 the whole process (this has already been discussed on p. 547). It is 

 evident that the metabolic rate declines with age, slowly before the 

 1 1 th day and afterwards much faster. If this graph be compared 

 with that already given for metabolic rate as calculated from the 

 gas and w^eight values of the Danish school (Fig. 142), it will be seen 

 that, although the metabolic rate in their case certainly declines, 

 it does so in harmony with the (Minot) growth-rate, being relatively 

 rapid at the beginning and relatively slow at the end. It therefore 

 does not obey Murray's rule. At present no explanation seems avail- 



