88 Prof. Magnus on the different Properties of the 



a rock-salt lens at the distance of its focal length, so that the 

 rays fell parallel upon the prism placed a short distance behind 

 it. Behind the prism was another lens, in the focus of which 

 the spectrum was formed. The spectrum was narrow, its visible 

 coloured portion having a width of only 4 or 5 millims. In order 

 to examine its various parts, the linear pile with a slit of 0*5 

 millim. was used. In other respects the same precautions were 

 adopted as in the experiments without lenses — the strips being 

 first compared before being platinized, one of them being plati- 

 nized after they had been found to be equal, and the effect of 

 both being then observed. 



The results that were obtained are represented graphically in 

 the same manner as the preceding ones. Figs. 2 to 5 show some 

 of the curves obtained. In constructing them, the deflections of 

 the galvanometer read off in millimetres, reduced to two- thirds 

 of their length, are laid down as ordinates. The abscissae are 

 four times as great as the distances between the several positions 

 of the pile in the spectrum*. 



In consequence of the flame radiating heat, as well as the 

 strips, the effect which it alone produced was determined a few 

 times, the bright and the platinized platinum and the flame alone 

 being successively allowed to act. In fig. 2 the smallest ordi- 

 nates represent the action of the flame alone. If these are de- 

 ducted from the ordinates which denote the effect of the flame 

 and the bright platinum, and of the flame and platinized plati- 

 num, the difference in the action of the two platinum surfaces 

 is seen to be still greater. Fig. 6 represents the curves of 

 these differences as they result from the values given in fig. 2. 



Conclusions. 



All these curves show that the maximum heating effect falls 

 in the dark part of the spectrum, that is, beyond the red, and in 

 all of them it is nearly at the same point. This maximum is 

 about twice as great for the platinized plate as it is for the bright 

 one. In all cases the increase of radiation shown by the plati- 

 nized plate takes place principally in the non-luminous part of 

 the spectrum ; so far as the coloured portion of the spectrum 

 extends, the heating caused by the platinized strip is scarcely 

 greater than that produced by the bright strip. Consequently 

 the wave-lengths the intensity of which is increased by platini- 

 zing are chiefly those lying beyond the red. These wave-lengths 

 are also the most intense among those given out by bright 

 platinum, and indeed by glowing bodies in general. 



The superficial areas which, in fig. 6, are contained between the 



* In fig. 4 the word " Blue " has been added by mistake, and " Yellow " 

 and " Red " ought to be further to the left. 



