PAPERS OX BIOLOGY AND AGRICULTURE 173 



different strata by these methods, although the ratios 

 so obtained are probably in the right direction and of 

 the right order of magnitude. The method of collection 

 was probably most efficient with respect to the leaf stra- 

 tum, the soil stratum and the herb and shrub strata fol- 

 lowing in the order named. The numerical data obtained 

 from samples taken in the manner described above are 

 given in the accompanying table and illustrated graphi- 

 cally in Figure 1. 



Examining the curve representing the total popula- 

 tion in all strata we find a fall in the fourth week (the 

 week of July 25), corresponding to an increase in the 

 evaporating power of the air at that time. At the be- 

 ginning of the period of study the woods were rather dry 

 and the animal population was small. With the increase 

 in moisture content of the air (and of the soil-leaf strata) 

 during the week of August 1, there was a marked in- 

 crease in the size of the collections. The increased humi- 

 dity, however, does not account for the high maximum 

 during the fourteenth week (ending Oct. 3). A gradual 

 decline in temperature had been going on with a consid- 

 erable increase in mean variability due to lower tempera- 

 tures at night, so that while the initial increase was due 

 certainly to a return to moisture conditions nearer the 

 optimum for the species concerned, the sudden and great 

 increase in the number of organisms taken in the samp- 

 les was due almost entirely to the great autumnal migra- 

 tion toward places of hibernation. An analysis of the 

 samples taken at this time showed that the great maxi- 

 mum was caused mainly by a few species of beetles and 

 leaf -hoppers, all hibernating forms in course of migra- 

 tion from the forest margin to the leaf stratum of the 

 more protected portions of the woods. Following this 

 maximum the fall was again rapid as temperatures con- 

 tinued to fall (the first heavy frost was the night of Oct. 

 4). The peaks in the total curve after November 12, the 

 date of the first ice formation, were caused by samples 

 taken on warm days when insects partially emerged from 

 hibernation or at least approached near enough to the 

 surface to be collected. The spring maxima may be ex- 

 plained largely as the result of the reverse movement. 



