Steam under certain circumstances. 125 



becomes colourless at small thicknesses, and absolutely opake 

 at greater. 



(3.) The state of tension of the steam seems only to affect 

 the phajnomena so far as it renders the critical colorific stage 

 of condensation more or less completely observable. 



(4.) The absorptive action of steam on the spectrum is not 

 exerted in the same way as that of other gaseous coloured 

 bodies, such as nitrous acid gas, and iodine vapour. It cuts 

 offj however, totally the same part of the spectrum as nitrous 

 acid does. Its phaenomena perhaps have a greater analogy 

 to those of opalescence than any other. 



The effect of mere change of mechanical structure in alter- 

 ing the optical properties of bodies is a phsenomenon likely to 

 give important information, both as to the constitution of 

 matter and the constitution of light; and the present observa- 

 tions may perhaps be one day viewed as a contribution to- 

 wards a mechanical theory of vapour, including that most sin- 

 gular stage which intervenes between the gaseous and com- 

 pletely liquid form, and which is probably connected with the 

 mechanical suspension of clouds. It is at all events very im- 

 portant to know that a portion of watery vapour confined in 

 a close vessel, and subjected to change of temperature alone, 

 without chemical change, is capable of undergoing the altera- 

 tions of colour and transparency which have been adverted to. 

 The singular fact noticed by Sir D. Brewster in the case of 

 nitrous acid gas, the colour of which deepens to an intense 

 orange red by the simple application of heat, seems to be a 

 fact of this kind. 



I cannot doubt that the colour of watery vapour under cer- 

 tain circumstances is the principal or only cause of the red 

 colour observed in clouds. The very fact that that colour 

 chiefly appears in the presence of clouds, is a sufficient refu- 

 tation of the only explanation of the pha?nomena of sunset and 

 sunrise having the least plausibility, given by optical writers. 

 If the red light of the horizontal sky were simply complemen- 

 tary to the blue of a pure atmosphere, the sun ought to set red 

 in the clearest weather, and then most of all ; but experience 

 shows that a lurid sunrise or sunset is «toff?/s accompanied by 

 clouds, or diffused vapours, and in a great majority of cases 

 occurs when the changing state of previously transparent and 

 colourless vapour may be interred from the succeeding rain. 

 In like manner terrestrial lights seen at a distance grow red 

 and dim whqii the atmosphere is filled with vapour soon to 

 be precipitated. Analogy applied to the preceding observa- 

 tions would certainly conduct to a solution of such appear- 

 ances ; for I have remarked that the existence of vapour of 



