it that he based this adverse conclusion on his own work, 

 which included the "Turbinares, Staganopodes, Pygopodes, 

 and Longipennes," a group of families forming but a small 

 per cent, of the Class, and also taking in birds largely of 

 the lower levels. Furthermore, he based his conclusions 

 (seemingly) on the mean of the temperatures of the species 

 in the family, which seems untenable, if it be true (as he 

 points out) that there are differences of temperature 

 amongst the families of the orders, and the more untenable 

 if the species within the family also show different tempera- 

 tures. Simpson also based his conclusions on the average 

 temperatures of the twenty-four-hour period, which I feel 

 is not correct; unless I am much mistaken, man's normal 

 temperature is regarded as the average of the highest in 

 twenty-four hours, i. e., that taken between noon and six in 

 the afternoon, and such is the method I have utilized, when 

 possible, in my consideration of this phase of this study. I 

 am strongly of the opinion that graduations of temperature 

 occur primarily in the class, secondarily in the order, and 

 again in the family ; what occurs in the whole avian tree is 

 repeated in the secondary, tertiary, and even smaller 

 branches. 



Seeking to confirm these ideas on grounds other than 

 those of these meagre records, one may justifiably ask, are 

 there any other facts (or reasons) which can point to, or 

 explain, the assumed co-existence of "high temperatures" 

 in "high birds," and "low temperatures" in "low birds?" 



It is not possible here to go extensively into the question 

 of the physiology of animal heat, but it can be examined 

 briefly at one or two points where it applies to the ideas now 

 in hand. It has been shown (148) that, with all mammals 

 investigated up to date, each animal can go through only a 

 definite and fixed number of metabolic changes during its 

 period of post-embryonic development, a contention based 

 on the fact that all such mammals consume equal total num- 

 bers of calories per kilo of weight, from birth to maturity ; 

 there are further indications that this ratio may also obtain 

 through the whole period of life. These facts mean that 

 the large mammal, which in general lives longer than the 

 smaller one, goes through its chain of metabolic changes 

 slowly, taking a long time to do so, while a small mammal 

 does the reverse, using up its metabolic-change quota (which 

 is identical in both mammals) swiftly. The large mammal 

 lives long and slowly, the small one briefly and swiftly, a 

 difference in "swiftness of life." What is known as to a 

 correlation between size and longevity in mammals tends to 

 confirm this theory of the "swiftness of life," for in the 

 scale of size represented by the elephant, camel, dog, and 

 mouse, variations in size and longevity are found to go hand 



59 



