the temperature coefficient for anv given process varies wj 



temperature, so that any mi sooicept.ions which may hare arise"- 

 have Veer due more to lacV of completeness in presenting the 

 logical analyses of the matters in hand than the actual fail- 

 ure to appreciate the raricus points of this analysis. 



Tt. should therefore he emphasized that there is nothing 

 particularly new about the points brought out ir the preceding 

 paragraphs ; they are probably clear to everyone who has ser- 

 iously studied rate-temperature relations. But writers appear 

 to have been prone to look only at certsin aspects of this rather 

 complicated problem, and to neglect other aspects just as im- 

 portant. Sreat emphasis should be placed upon the fact that 

 the temperature coefficient for any process having temperature 

 limits is a continuously varying value, the variation proceed: 

 from infinity to zero. 



Prom this point of view the tsmperaturs relations of 

 different processes under-stated ncn-tamperature conditions and 

 with stated exposure periods are clearly comparahle, not by means 

 of single temoerature coefficient values, but by means of the 

 coefficient-temperature relations as a whole. 'Practically, the 

 simplest way to present this relation for a given process is 

 to construct such 3 coefficient-temperature rraph as those 

 shown in fi?-uB9 11. s form of this graph and its position 



within the anfle bet. veer the rectangular axes includes a com- 

 plete description cf the rate- temperature relation. If two pro- 

 cesses are to be compared in respect to this temperature rela- 

 tion t>e comparison should be instituted between the two co- 

 eff ici en t- temperature graphs, constructed on the same scale, 

 if the two graphs coincide throughout, then the * .-rperati.tr 3 



