PARTICLES IN FOGS AND CLOUDS. 419 



thick form ; and in the experiment above described, the lamp products, when expanded 

 quickly, were many times thicker than when they were expanded slowly. All this 

 greater density when the expansion was quick was not, however, due to the greater rate 

 of condensation. When the air is expanded slowly, there is time for it to receive heat 

 from the walls of the receiver and by radiation ; the air is therefore not cooled so much 

 when expanded slowly as when quickly. But if we vary the experiment in the following 

 way, we shall find, that for the same amount of cooling, quick expansion does really give 

 the densest condensation. Connect one of the glass receivers with the metal vacuum 

 receiver, and also provide the glass receiver with a vacuum gauge — a long glass tube with 

 its lower end dipped in water will do. Now make a slow expansion of the air in the glass 

 receiver, which can easily be done by opening the stopcock of the metal receiver very slowly 

 and regulating the opening by observing the rate of the rise of the water in the vacuum 

 gauge. Note the maximum density of the clouding in the receiver, and note also the highest 

 point to which the water rose in the tube when the expanding was stopped. Now ob- 

 serve the amount the water falls in the vacuum gauge after the expansion is stopped, due 

 to the air in the receiver recovering its original temperature. The amount of this fall gives 

 the cooling effect of the slow expansion. Now arrange matters so as to give a quick 

 expansion which will have the same cooling effect. The amount of expansion will now 

 be very much less than when the expansion was slow, but it will be noticed that the 

 density of the condensation is much greater. It is, therefore, evident that the greater 

 thickness obtained when the lamp products were expanded quickly, was greatly due to 

 the rate at which the condensation was made to take place. 



Further evidence on this point has already been given in the paper referred to " On 

 some Phenomena connected with Cloudy Condensation." It is there shown that when 

 steam is made to issue under pressure from small nozzles, or from rough nozzles, or in any 

 way made to mix quickly with the cold air, the condensation is far denser than when 

 allowed to mix slowly. 



There are some further precautions necessary in interpreting the experiment with the 

 products of combustion from the lamp and from the gas flame in which a little sulphur 

 was burned. Both fogs cleared away, but the fog made in the lamp products cleared 

 very much quicker than the other. Now, in both cases, only part of this clearing was 

 due to differentiation and falling ; part was due to rise of temperature. The air tended 

 to regain its original temperature after the expansion ; the drops were therefore partly 

 evaporated. It is, however, found that though the air is prevented from regaining its 

 original temperature, all the phenomena are much the same. If, for instance, we cool 

 the walls of the receiver with snow just a moment before the expansion is made, though 

 the air cannot under these conditions regain its original temperature, yet the condensa- 

 tion rapidly clears very much as it did when the receiver was not cooled. Again, if we 

 use a jacketed receiver, and keep the jacket at, say, 32°, and rapidly admit air at 60°, 

 and expand it before it has time to be cooled by the receiver, so as to have it surrounded 



