of dark Heat-rays by Gases and Vapours. 5 



Absorption per 100*. 



, A x Absorption in 



Blackened tube. Polished tube. blackened tube 



Vapour. Pressure 03 inch. Pressure 5 inch, proportional to 



Sulphide of carbon... 5 27 f 23 



Iodide of methyl 15*8 60 71 



Benzol 17*5 78 79 



Chloroform 17-5 89 79 



Iodide of ethyl 21*5 94 97 



Wood-spirit 26-5 123 120 



Methyl alcohol 29-0 133 131 



Chloride of amyl 30'0 137 135 



Amylene 31"8 157 143 



The last column gives the values of the first column multi- 

 plied by 4*5. Of these nine substances, we find three for 

 which the results deviate considerably from proportionality. 

 Thus, the number for iodide of methyl in the last column dif- 

 fers by 16 per cent, from that in the middle column, the num- 

 bers for chloroform by 11 per cent., and for amylene by 9 per 

 cent. If three out of nine substances (that is, one third) do 

 not show this proportionality, it is to be expected that, by in- 

 creasing the number of substances experimented upon, other 

 deviations would be brought to lio-ht — deviations which mio'ht 

 make it impossible to maintain the theory of proportionality. 

 Tvndall himself singles out the following cases: — 

 " With the blackened tube the order of the following sub- 

 stances, beginning with the lowest, was 



Alcohol, Sulphuric ether, Formic ether, Ethyl propionate, 



whereas with the polished tube the order was 



Formic ether, Alcohol, Ethyl propionate, Sulphuric ether " 



— thus altogether different ! But we find in other tables values 

 for the absorption of these substances in a polished tube at 

 0'5 inch pressure as follows: — 



Formic ether (Table XIV. p. 32) ... 133 



Alcohol (Table VII. p. 33) 157$ 



Ethyl propionate (Table XV. p. 32) . . 168 



Sulphuric ether (interpolated) . . . . 180 



possible [The reader is requested to bear in mind that, although in 

 " ", the words used are not precisely those employed by Dr. Tyndall, but 

 a literal translation of the German version. — Eds. Phil. Mag-.] 



* The numbers under " absorption per 100 " are obtained from the de- 

 flection of the galvanometer, a deflection of 1° being taken as unit. Else- 

 where " absorption per 100 " denotes the percentage of the total radiation. 



f According to Table VI. p. 27, we should have expected 14-8 instead of 

 this number. 



\ At the place cited the number is given as 175, which must certainly 



