148 INFLUENCE OF TEMPERATURE ON BIOLOGICAL SYSTEMS 



to be at 4°, 8° and 13°C, respectively. Similar thresholds can be defined 

 for Oncopeltus where development to blastoderm occurs at 5°C, develop- 

 ment to a complete nymph requires a little above 10°C, hatching requires 

 13°C, complete development plus hatching requires 15°C, and full viability 

 of the hatched bug requires something above 17°C (5). A longer series of 

 thresholds could be defined if one wished to do the necessary work. The 

 first threshold which will be dealt with here is the 'developmental-hatching 

 threshold' of Johnson but for simplicity of expression it will henceforth be 

 called the 'hatching threshold.' 



One of the most serious limitations to series of thresholds, such as those 

 above, is that they are determined on a basis of incubation at constant 

 temperatures or (for the 'hatching threshold' of Johnson) incubation at 

 one constant temperature followed by a single shift to another constant 

 temperature. Determinations of threshold values are repeatable only under 

 identical conditions. Thus we find that if one uses varying temperatures 

 instead of constant temperatures (such as several hours at 20° or 25° 

 followed by the remainder of each day at 13° or 14°) the thresholds when 

 stated as average temperatures are found to be 1-2°C lower (8) . Another 

 limitation we find is that the hatching temperature threshold is not in- 

 dependent of all other factors. If the relative humidity is lowered from 

 the optimal value of about 75% to 50%, hatching is poorer or, otherwise 

 stated, the temperature for the hatching threshold is raised several degrees 

 (8). One may object to comparing the effects of an average temperature 

 with those of its corresponding constant temperature, or to modifying 

 other environmental conditions, but clearly the thresholds are difficult to 

 define satisfactorily and can be defined only in terms of a fully stated set 

 of conditions. 



Incidentally, no one of these thresholds in Oncopeltus corresponds to the 

 chill-coma temperature, which for just-hatched bugs is 3°, or for bugs a 

 day or two after hatching is 2°C. 



Confining our attention for the moment to the hatching threshold [i.e. 

 that temperature which when maintained constantly with optimal hu- 

 midity will just permit full embryonic development plus hatching) we 

 start by acknowledging the correctness of the day-degree accumulation 

 idea. The developing embryo acts like an adding machine. It summates rate 

 at particular temperatures X time at those temperatures, and hatches when 

 the calculated 100% of developmental time has been attained (8). Dis- 

 crepancies occur in alternating temperature experiments, especially when 

 one of the temperatures is subthreshold. A large part of the larger errors, 

 however, is readily ascribablc to uncertainty about the determination of 

 rate values to use for subthreshold temperatures (shown by the >10% 



