It is obvious that the temperature coefficient of the rate of oxidations is remarkably constant, about 2 for 10°, for various temperatures and does not show the variation from 7 or more to 2.2 for Q{ 10} for the rate of segmentation.
Kanitz[256] has shown that in a graph in which the logarithms of the segmentation velocities are drawn as ordinates and the temperatures as abscissæ the logarithms form two straight lines which are joined at an angle. According to the law of van't Hoff and Arrhenius concerning the influence of temperature upon velocities of chemical reactions the logarithms should lie in a straight line. We are dealing therefore in these cases with two exponential curves, one representing in _Arbacia_ the interval 7-13° and the second from 13-26°; in _Strongylocentrotus_ between 3-9° and 9-20°.
It was found in these experiments that if measurements of the Q{ 10} of later stages of development are attempted the variations due to unavoidable difficulties become too great to permit an equal degree of reliability in the determinations.
The vast importance of this influence of temperature upon the rate of development is seen in the fact that in addition to the food supply the rate of the maturing of plants and animals depends on this factor.
2. This influence of temperature upon development has been used to find the conditions determining fluctuating variation. The reader knows that by this expression are understood the differences between individuals of a pure strain or breed. These variations are not inherited, a fact contrary to the idea of Darwin, who assumed that by the selection of extreme cases of fluctuating variation new varieties could develop. What is the basis of this fluctuating variation? The writer concluded that if fluctuating variations were due to a slight variation in the quantity of a specific substance--in some cases an enzyme--required for the formation of a hereditary character, the temperature coefficient might be used to test the idea. We have just seen that the time required from insemination until the cell division of the first egg occurs is very sharply defined for each temperature. If a large number _e. g._ one hundred or more eggs are under observation simultaneously in a microscopic field it can be seen that they do not all segment at the same time but in succession; this is the expression of fluctuating variation. Miss Chamberlain and the writer have measured the time which elapses between the moment the first egg of such a group segments and the moment the last egg begins its segmentation, and found that this latitude of variation is also very definite for each temperature, and that its temperature coefficient is for each interval of 10° practically identical with the temperature coefficient of the segmentation for the same interval.[257] The slight deviations are practically all in the same sense and accounted for by a slight deficiency in the nature of the experiments. The two following tables give the latitude of variations for different temperatures for the first segmentation in _Arbacia_ and the temperature coefficient for this latitude and the rate of segmentation. These two latter coefficients are practically identical.