84 Prof. Magnus on the different Properties of the 



radiation to double its previous amount. In the next place, 

 plates of various substances were interposed by fastening them 

 in front of the diaphragm mentioned above, and the deflections of 

 the galvanometer were observed, which were produced by the 

 rays transmitted by these substances, both when the source of 

 heat was the bright disk and when it was the one covered with 

 platinum-sponge, which, for the sake of shortness, I shall call in 

 the sequel the platinized disk. 



The substances of which plates were interposed were rock- 

 salt, calc-spar, rock-crystal, smoky topaz, agate, plate-glass, flint- 

 glass, and dark green glass, each from 6 to 7 millims. thick. 

 Red, orange, yellow, green, blue, violet, as well as polished and 

 rough colourless glass, were likewise used, the thickness of each 

 of these being about 2 millims. . ,, . 



A number of these plates absorbed proportional quantities of 

 the heat from both sources ; so that if I denotes the deflection of 

 the galvanometer caused by the direct radiation of the bright 

 disk, and i the deflection produced when one of the above-men- 

 tioned plates was interposed, and Ij and i x denote the corre- 

 sponding deflections when the platinized plate is used, we have 



i __ i x 



1 7" IT* 



A short time ago Professor Tyndall published the interesting 

 observation that sulphide of carbon, when rendered quite opake 

 by dissolving iodine in it, transmits heat quite as well as it does 

 when pure and transparent. It appeared to me that it would 

 be interesting to interpose these two liquids, namely transparent 

 sulphide of carbon, and the same substance made opake by 

 iodine. For this purpose, layers of them 10 millims. thick were 

 enclosed between two very thin plates of rock-salt. But these 

 two liquids also absorbed the heat from both sources in the 



i 1 



same proportion, so that here again we had y = ~- 



i 

 With a number of the other plates above mentioned, j was 



always found greater than y- 1 . The difference varied with dif- 

 ferent plates, but with none of them was it very considerable. 



ner : — A little chloroplatinate of ammonium is stirred up with water to a 

 thin mud, and spread upon the plate in a very thin uniform layer with a 

 hair pencil ; the plate is carefully dried and heated in a flame ; a second 

 coating is then applied in the same way, and this is repeated till the re- 

 quired thickness is attained. If the chloroplatinate of ammonium is ap- 

 plied in a thick layer to begin with, it collects together in separate masses 

 on heating, and so does not cover the plate uniformly. 



