20 Ml! JOHN AITKKX ON 



when tested iu the flask, I, but the density is not very much greater than what the air 

 of the room would give if it had not passed through the hydrogen flame. The concen- 

 trated light from the lantern produces a luminous cone shining in the flask. If now, 

 however, we move about things in the room, so as to stir up some dust, the result is 

 remarkable. On now testing the products from the hydrogen flame in the flask the 

 cone of light is no longer visible. So dense is the condensation that the whole flask 

 is filled with a dense white fog, the flask looking like a white ball, the intense heat of 

 the hydrogen flame having produced an enormous number of nuclei out of some of 

 the dust floating in the air. 



The conditions of the experiment, in which it is shown that hydrogen burning in 

 filtered air gives rise to no nuclei of condensation, is not, however, quite the same as that 

 in which Helmholtz found these products to be very active, as he mixed them much 

 more quickly with the re-acting agent. The products in his experiment were carried 

 direct to the steam jet. To test the condensing power of the products immediately after 

 combustion, the following alterations were made in the arrangement already described in 

 the apparatus shown in the Plate : — The test flask, U, was introduced into the circulating 

 .-ystem at a point close to the combustion chamber, so that the products might be tested 

 immediately after combustion. For testing them in this case the steam jet method was 

 employed as in Helmholtz' experiments, only no unfiftered air was allowed to mix with 

 the products. Steam was generated in the flask, W, and conveyed by means of a glass 

 tube to the interior of the test-flask, Z7. This tube terminated in a fine jet. With the 

 hydrogen burning and unfiltered air, there was a dense condensation, but when filtered 

 air was used the condensation was extremely slight, and if the flame was low on some 

 trials, it was difficult to say whether there was any condensation or not. In each test 

 the steam was allowed to enter only for a very short time, as the test-flask got heated, 

 and had, therefore, to be cooled before each test. It is evident from these experi- 

 ments that the products of combustion from pure hydrogen and dustless air are far 

 from being very active in the condensation of the steam jet, even when they are 

 newly formed. 



There are other reasons for leading us to conclude that the action of these "ions" is 

 extremely short-lived. It is generally admitted that these "ions" are the cause of 

 electrical conduction in gases, that it is to their presence in the products of combustion 

 that these gases owe their power of discharging electrified bodies. Now we know that 

 the products of combustion soon lose this power; if we bring the products from the 

 flame to the electrified body by means of a short tube, the discharge is rapid, but as 

 we lengthen the tube, the power very rapidly decreases. Again, the air of a room 

 where gas is burning has very little, if any, power of discharging electrified bodies, 

 yet it has enormous numbers of nuclei of condensation, due to the products of 

 combustion. Lord Kelvin and Dr Magnus Maclean, in a recent communication to 

 tin Royal Society of Edinburgh, show that the products of combustion lose their 

 power of discharging bodies in less than fifteen seconds. Further, we know that air 



