198 Annals Entomological Society of America . [Vol. XIII, 



humidity, multiplication is very slow and breeding experiments 

 usually fail at a temperature of 70° F. (a much higher limit 

 than that of any other insect studied), but that with high 

 humidity and high but variable temperatures like those of a 

 hot, moist day in summer, breeding experiments are highly 

 successful, and the rate of multiplication is almost unbelievably 

 rapid. 



I am beginning to hope that, by vivarium work and com- 

 panion studies in the field, we shall be able soon to standardize 

 our life history data so that we can describe the life cycles of 

 insects, not primarily in unreliable units of time, so variable as 

 to be perplexing, but in ecological units of temperature, humidity 

 and the like, invariable for a species whenever and wherever it 

 may be found. An example of such a standardization is 

 furnished by the product of recent work on the life cycle of 

 the codling-moth done by Mr. Glenn at a well-equipped orchard 

 station in southern Illinois, to the effect that in normal seasons, 

 when the sum of all mean daily temperatures which fall between 

 50° and 85° F. reaches 550°, the eggs of the spring generation of 

 the moths will begin to hatch, and when these totals reach 

 1550° those of the second generation will hatch, and 

 when eleven hundred degrees is added to this sum, the third 

 generation may begin to hatch; but that if this last total of 

 2650° is not reached before September 10, there will be no third 

 generation at all in that year; and these statements may be 

 expected to hold good every year without regard to the character 

 of the season or to differences of elevation or latitude. Dr. 

 Shelford is now working out in the vivarium a scheme of cor- 

 rections to be applied to this forecast whenever the humidity 

 factor is practically important. 



Our present method of describing life histories in days or 

 hours for each stage or phase is, indeed, thoroughly illogical, 

 for mere lapse of time has, of course, no effect in itself; it is 

 only the dynamic content or accompaniment of the time unit, 

 especially in temperature and humidity, which really signifies. 

 We must find our unvarying ecological constants and make up 

 our life-history calendars of these and not of the uncertain 

 units of time which we now use simply because they are the 

 most easily obtainable. 



Perhaps we shall never know just how and to what immedi- 

 ate profit the holometabola were differentiated, but that the 



