spaces, and even possibly by geologic time-spaces; thus, the 

 swallow and the finches of Australia still cleave to the an- 

 cestral period of their cousins of the North, and the English 

 and New Zealand gannets have identical incubation periods. 



Certainly the incubation length seems far more change- 

 less, persistent and deep-rooted in natiu'e than are other 

 characteristics of birds, as, for example, the proventriculus 

 mucosa of a gull (25). Furthermore, as has been suggested, 

 it is not subject to selection by man, as are other character- 

 istics, for while he has, in effect, changed a jungle fowl into a 

 five-toed dorking, and a rock-pigeon into a fan-tail, still 

 both of these species have retained their original lengths of 

 incubation. 



If the structure of the egg shell persistently tend to be 

 characteristic, and remain the same with groups of birds, or 

 with the species (110), why should not the far more im- 

 portant process of embryonic development, and its length, 

 do likewise? While the immediately foregoing would seem 

 to show the length of the i:)eriod of incubation as a fixed 

 characteristic, yet a superficial examination of a list of such 

 periods leads one to believe that there is, in fact, considerable 

 variability in the length of incubation of a given species, 

 and unless one can learn if this be true or false, it were 

 absolutely useless to attempt to draw conclusions from the 

 facts published on incubation lengths, because, viewing these 

 facts as evidence, they are in many parts hopelessly con- 

 flicting. I am convinced that most of such conflicts are to 

 be explained by a careful study of the influences apparently 

 affecting the true length of incubation, /. e., influences caus- 

 ing apparent variability in the true length of incubation. 



Variability in the length of incubation may be tTue, 

 that is, permanently lengthening, or actually shortening, the 

 minimum number of days for successful hatching under 

 optimum conditions, and apparent, shortening or lengthen- 

 ing by slowing of the embryonic development by errors of 

 faulty time measurements. 



True variahility — The writer questions very much 

 whether there be any decided trve variability, i. e., a vari- 

 ability occurring when all necessary conditions are optimum. 

 There is a small amount of experimental evidence at hand 

 which shows conclusively that with the domestic hen it is 

 possible, by suitable regulation of temperature conditions 

 in an artificial incubator (33), to shorten the length of in- 

 cubation a few hours only. This is well known to poultry 

 raisers, who know also that the dividing line between suc- 

 cessfully shortening the period and killing the embryo is 

 exceedingly difficult to maintain, even impossible at times. 

 All secondary influences which tend to induce this subtle 

 influence of temperature increase, especially towards the 



16 



