1. Diurnal fluctuations in emergence of the adults and negative correlation between the length of successive stages in the puparium made it desirable to restrict study of relation of temperature to development to the prepupal stage.
2. On morphological grounds, the formation of the puparium, which starts the prepupal period, seemed to be determined by the stage of larval development; pupation, which terminates the prepupal stage, by imaginal disk development.
3. The rate of prepupal development may be represented by the Arrhenius equation relating velocity of an irreversible chemical reaction with temperature. The data gave three values for the critical increment over different temperature intervals, corresponding to three straight lines of different slope. When deviations of the points from these lines were compared with their probable errors, however, in nearly three-fourths of the cases the difference was significant.
4. Analysis of these deviations showed them to be due primarily to changes in the extent of imaginal disk development at the time of puparium formation. These, in turn, were correlated with age of the culture.
5. The two sexes differed in developmental velocity, such that the rate of female development was about 1.03 times as great as rate of male development. For the upper temperatures this ratio was greater than for the lowest of the three temperature ranges, the intermediate zone possibly varying between the two.
6. A final curve relating prepupal development to temperature has been calculated after (1) converting all female records to a male basis, (2) applying a correction for age of culture error, and (3) weighting each point by the square root of the number of cases upon which it was based. This yielded the following values for the temperature characteristic µ, namely, 33,210 from 12–16°, 16,850 from 16–25°, and 7,100 from 25–30°. Above 30° the data could not be fitted by the Arrhenius equation.
7. By treating prepupæ in different developmental stages to brief exposures at a lower temperature, pupation was more delayed by treatment at the beginning than at the end of the prepupal stage. From these data, µ for parts of the stage were calculated on the assumption that the effect of temperature did not persist after return to the standard temperature. Since the µ thus secured were greatest for the beginning and least for the end, and none were less than that for the whole, the interrelations of the successive stages are probably more complex than they were assumed to be in making the calculations.
8. Lowering the temperature prior to puparium formation lengthened the prepupal stage. Puparium formation, therefore, was not conditioned by imaginal disk development, but by larval processes possessing a lower temperature coefficient than did the imaginal disks.
9. Possible physicochemical mechanisms for producing these results are discussed, and the observed temperature characteristics were finally attributed to three relatively independent anabolic processes which limit the rate of cell growth in the imaginal disks.
10. Both the thermal increments and the critical temperatures for the prepupal stage differed markedly from those reported by Loeb and Northrop for the entire development within the puparium. Since the prepupa forms part of the latter period, temperature characteristics for extended developmental phases known to be heterogeneous are of doubtful significance.