432 PROCESSES INFERRED FROM INDIRECT OBSERVATION 



pairs of distances and they differed more or less from the ideal value 

 of unity, but the average of a large number of these ratios differed from 

 unity only by an amount commensurable with the probable error of 

 the average. The following are their results, the values enclosed in 

 brackets being rejected by Chauvenet's criterion for the rejection of 

 extreme variates. 1 



* The averages and probable errors given are those recalculated by the authors 

 since the original article was published. 



It is a general law of Photochemical Action that only those rays are 

 effective which are absorbed by the system in which the reaction 

 occurs. Visible light-rays are not, as a general rule, selectively ab- 

 sorbed by protoplasm and hence their action is usually confined to or 

 exerted reflexly through specialized pigmented areas which constitute 

 the receptive elements of optical sense-organs. White light which is 

 not toxic for the majority of tissues may be rendered toxic, as L. Loeb 

 has shown, by impregnating the tissue with certain dyes, particularly 

 Eosin, which in such cases acts as the photochemical absorbent or 

 sensitizer. Ultraviolet Light, however, is universally toxic even for 



1 W. Chauvenet: A Manual of Spherical and Practical Astronomy, Philadelphia, 

 1891, vol. 2, p. 558. 



