4 FREER. 



would have increased the intensity of illumination, or the amplitude of 

 the light waves, a process corresponding to an increase of voltage, with 

 diminution of the diameter of the cross section of the wire, in an 

 electric conductor. While it is easy to conceive of measurements of 

 this kind where rays pass through a definite, limited opening, it is not 

 so easy to see how they can be made when the illumination extends over 

 a vast area. Bunsen and Eoscoe 3 and Eoscoe and Baxendell,' 1 in their 

 classical researches on photochemistry, employed an explosive mixture of 

 chlorine and hydrogen to measure the intensity of insolation, and by a 

 variation in the quantity of light, by using different sources and different 

 diameters of openings, came to a series of results which, more nearly than 

 any other, can be used as a basis of calculation. However, it seems to me 

 that the possibilities of autocatalysis in such a mixture, 5 which have 

 since then become understood, would have a great influence in the cal- 

 culation of the total effect of insolation in climates other than the one 

 in which these investigators worked. Subsequently, they also perfected 

 a method based upon the darkening of a normal strip of sensitized paper 

 in known intervals of time to standard colors, and this means, if the 

 paper were always correct, could be developed in the direction mentioned. 

 Indeed, Eoscoe ° described a self-registering apparatus for use in me- 

 teorological stations which would, if universally adopted, already have 

 made available the data sought for. However, it seems to me that all 

 methods which consider "photochemical" light only as a group, without 

 division by the spectrum, are faulty because they do not differentiate 

 the proportion of waves of various lengths contained in the total insola- 

 tion at different latitudes and under different meteorological conditions. 

 The range of rays which will bring about the union of hydrogen and 

 chlorine, or the darkening of sensitized paper, is very great, whereas it 

 may be only certain very limited portions of the spectral field which 

 produce maximum results. 



One method suggests itself. If two spectra of the sun, each showing 

 the same area, that is, composed of waves of the same length, were to be 

 projected upon two extremely slow photographic plates, so slow that. 

 although the plates are approximately the same, slight variations in time 

 would have no appreciable effect upon the total exposure, and if compari- 

 sons were to be made as to the time necessary to produce the first visible 

 image of some given line, then the ratios of these times would, for most 

 practical purposes of comparison, give the ratios of intensities. The 

 accuracy would increase by selecting more than one line, successively. 



'Lor. cit., ibid. (1862) 117, 529. 



'Ann. d. Phys. it. Chem. (Poggendorf) (1866) 128, 291. 



'Chapman, Chadwick and Ramsbottom. Journ. Chem. Soc. Loudon (1907) 91, 

 943, give a parallel ease in the reaction of carbon monoxide and oxygen in the 

 presence of moisture. Hydrochloric acid takes the place of the latter. 



'Ibid (1877) 151, 268. 



