and birds). Therefore, in this discussion, a bird's normal 

 temperature will be taken to mean, not the mean of the 

 records of twenty-four hours, but the highest in that period, 

 or the average of such highest records ; hence, the most desir- 

 able bird temperature records are those taken between noon 

 and six in the evening, accompanied by notes as to the bird's 

 sex, age, method of securing the bird, and of other modify- 

 ing factors, some of which have just been outlined. 



The way of securing a bird to take its temperature 

 varies, and it may have considerable influence on the result. 

 Sea birds were caught with a baited hook and line for such 

 purposes, nearly a century ago (165). I have found, using 

 trapped English sparrows, that the rectal temperature of 

 these birds is the same before and directly after being shot, 

 and Simpson (165), in studying other species, antedated me 

 in this conclusion more than thirteen years. This conclusion 

 is also true of flickers, since a male's temperature, directly 

 after being shot, was 106.6° F., while that of a female, taken 

 alive (and afterwards liberated), was 106.4° F. 



This method of securing animal temperatures is approx- 

 imately accurate, and is substantiated by similar methods 

 with mammals (156) ; in all cases, care must be taken to note 

 if the specimen bleed profusely (becomes exsanguinated), 

 in which case there would be a swift descent of the tem- 

 perature, or if the brain be extensively damaged, in which 

 event it is possible that the temperature might be, for a 

 short time, abnormally high. Whatever the method of se- 

 curing a bird, a standard self-registering clinical thermom- 

 eter should be used, inserting it into the bowel a half-inch 

 or more, according to the bird's size, and held in place until 

 the mercury ceases to rise. 



It seems self-evident that the ideal. time to secure bird 

 temperature records would be while a bird is inciibating, but 

 this, at present, would be difficult to do to any extent large 

 enough to be useful, and the next best time to take birds' 

 temperatures would be while they they were breeding or in 

 the brooding period. For the purposes of this study, an\ 

 and all recorded bird temperatures must be utilized, even 

 if all have been taken with little or no regard to their bear- 

 ing on the length of incubation or to their relation to the 

 peculiarities of the daily temperature curve. The number 

 of published temperature records of birds is not iarge ; aside 

 from some records scattered through the literature of human 

 and comparative physiology, and a few determined by my- 

 self, there are two principal sources of information on this 

 subject, the first being found in a list given by H. Milne- 

 Edwards (38), in 1863, and the second, a brief but highly 

 suggestive study published by Sutherland in 1899 (112). 



51 



