trying to solve the riddle of the cause beneath the varying 

 lengths of incubation amongst birds. 



Any attempt to draw conclusions from lists of incuba- 

 tion periods heretofore published seems, on preliminary 

 examination, to be hopeless ; the evidence in places is highly 

 conflicting and inconsistent, so much so that one is at once 

 tempted to believe the length of the incubation period is a 

 matter of more or less chance, and controlled by no particu- 

 lar condition or set of conditions. Fortunately, however, 

 years of observations on domesticated birds and a vast 

 experience in the use of artificial incubators show that this 

 conception cannot be true, and also show that there is an 

 actual, or a relative, fixity of the length of incubation with 

 such species as have been so domesticated. Furthermore, 

 the evidence seems to show that there is no inherent or 

 known reason why a similar specific fixity should not apply 

 to all avian species. 



It is regretted that the writer did not have personal 

 access to a larger mass of literature, for such would probably 

 have yielded many more records of incubation, additions 

 which would have greatly enhanced the value of these pres- 

 ent data, and would also have saved future students of the 

 same problem much drudgery in a search for such data. Of 

 the nineteen thousand or more (138) species and subspecies 

 of birds now known to ornithologists, the six hundred and 

 twenty-five species and subspecies given in this study form 

 but a small per cent., which may in fact be too small on 

 which to safely base final conclusions. A future larger and 

 more comprehensive study of the question will alone decide 

 this. 



The conclusions in this study are based on the assump- 

 tion that artificial incubator records, and such other records 

 as show a substantial concordance, are correct, and hence 

 justifiably available as fundamental data. 

 Conflicts in the Data 



It seems appropriate here to consider the fact that there 

 are many conflicts in the published records of incubation 

 lengths, and when I have been unable to examine the orig- 

 inal record, it has been accepted as quoted. When a record 

 has been secured by me from an original source, it has been 

 copied verhatim^ excepting in a few instances where it was 

 perfectly obvious from the context that the period had been 

 incorrectly estimated because of errors induced by "succes- 

 sive hatching," and in such a case the writer has tried, care- 

 fully and impartially, to make corrections for such errors, 

 and has listed the record as so corrected. 



The conflicts in the records given for a single species 

 are often numerous, and are accountable, many times, by 

 errors brought about by the difficulty of measuring the in- 



11 



