﻿88 Experiments with Carbon Dioxide in the Solid State, 



and working also with 3 grammes of solid matter, I now 

 collected only 150 cub. centim., and carbon dioxide in gaseous 

 condition is considered a bad conductor of beat. But the 

 difference in quantity of gas evolved depends also on the 

 power for convection, which will not be equal in the three 

 cases *. 



The following experiment, which is somewhat the reverse 

 of the former, demonstrates the influence of convection in 

 incandescent electric lamps in a striking way. Four similar 

 lamps (16 candle-power and same voltage) were connected 

 in parallel to a dynamo ; the first was filled with gaseous 

 carbon dioxide, the second with common coal-gas, the third 

 with hydrogen, and the fourth was kept in normal condi- 

 tion, that is provided with its vacuum around the carbon 

 filament. All carried a little piece of phosphorus at the top 

 of the glass globe on the outside, and now admitting the 

 current in all at the same time, it is observed that the 

 phosphorus is set on fire at different rates; first it burns on 

 the lamp with carbon dioxide, then on that containing coal- 

 gas, and shortly after on the hydrogen lamp, but on the 

 vacuum lamp it remains for a long time intact. In the latter 

 case it may be inferred that the dark heat-rays are very 

 imperfectly transmitted by the vacuum, whilst convection of 

 course must also be very reduced; hence the glass can only 

 receive a small amount of heat, and as in Prof. Dewar's 

 experiment the vacuum prevents the ready access of heat 

 to the extremely cold liquid, in the incandescent lamp, 

 on the contrary, it is an obstacle to the cooling of the filament 

 of carbon. It can therefore attain a high temperature and 

 convert the electric energy it receives, chiefly into radiant 

 light. As for the other lamps, the difference in heat trans- 

 ferred in the same time outside may find its explanation in 

 the well-known experiments of Grove t and the investigation 

 of Clausius; yet it is a curious fact to see the carbon brightly 



* Prof. Kundt described, as long ago as 1877, an experiment of a similar 

 character, employing three vessels of the same size, enclosing at a 

 distance of 3 mm. little tubes each filled with the same volume of 

 sulphuric ether. The vessels were filled respectively with hydrogen, air, 

 and carbon dioxide acting as jackets. Putting all three together in 

 boiling water, heat is transmitted in a different way, as appears when the 

 ether vapour is lighted, and the flame of hydrogen is seen to be the 

 longest, that of the carbon dioxide the smallest, but convection may also 

 influence. 



t Grove published his paper on this subject in the Phil. Trans, in 

 1847, and he determined the amount of heat given off by a platinum wire 

 in different gases to a surrounding mass of water ; the experiments with 

 the lamps are in accordance with his results. 



