

Lent for any tempera* >r • decreases i ritude, from time 

 period to time period for a certain time after inoculation, 

 after which tl e nagnitude rejnains approximate] ! same, cs- 

 cil&ting slightly, perhaps within the limits of what may 

 called observational error. 



The relation of the value of this temperature coef - 

 cient to the maintained temperature rep r eser, ting the middle 

 point of the 10-degree temperature range froir which each coef- 

 ficient value is derived, is shown graphically, for the second 

 24-hour period after inoculation, for each funp-us, in figure "4* 

 Abscissas are the temperatures of these middle points, while 

 crdinates are coefficient values. These graphs have not 

 ed. 



These four graphs of 10-degree temperature coefficients 

 are seen to he alilce in their general form. t apins 



With a relatively - Lgh value at the left (lowest tempera- 



e range tested) and descends, rapidly at first and V- 

 less rapidly, with high,- 3. 



■ 3 nature of the temperature coefficient it is 

 clear tha + the val e of any range of maintained es 



having its lower limit just below the mini mum temperature f or 

 enlar ge ment must he infinite ; 

 is e 1 Bed to vary from any value below the minimum temj 



re to any value above the minimum, the rate of ei La 

 must (by definition) suddenly ire-ease from zerc . s posi- 

 tive value, and the ratio of any posi*. .uanti • i;ero 

 of course infinil . All of the graphs should the 

 vertical Btraight lines for .hose 



