420 Prof. Rijke on a New Method of causing a Vibration 



distance of the disc from the lower end of the tube should be 

 one-fourth of its entire length . 



6. As to the explanation of the phjenomenon, that, I think, is 

 not hard to find. In warming the wire-gauze, the temperature 

 of the sides of the tube is raised also. If the lamp be then with- 

 drawn, an ascending current of air is established, which, passing 

 through the meshes of the wire-gauze, is necessarily heated, and 

 in consequence dilates. To this dilatation immediately succeeds 

 a contraction, due to the cooling effect of the sides of the tube. 

 It is to these successive dilatations and contractions, in my 

 opinion, that we must attribute the production of the sounds the 

 origin of which is in question. 



It is obvious that the presence of the heated disc alone would 

 suffice to produce an ascending current of air, and that the eleva- 

 tion of the temperature of the sides of the tube is rather injurious 

 than otherwise to the success of the experiment. In fact, that 

 which is of the greatest importance is, that the diffei'ence of tem- 

 perature between the air which rises in the tube and the meshes of 

 the wire-gauze should be the greatest possible. Several considera- 

 tions support this view of the case. For instance, we have seen 

 (1) that the tube only began to sound after a lapse of some 

 instants from the time when the lamp x^'tis extinguished ; now it 

 is evident that, during this interval, the air warmed by the 

 flame of the lamp, as well as the products of combustion, must 

 have been replaced by a column of air of a much lower tem- 

 perature. If the sound only lasts some seconds, that is because 

 the gauze is rapidly cooled by the ascending column of air ; and 

 if an augmented number of discs gives an increased duration to 

 the sound, that is because the presence of a greater number of 

 discs, by diminishing the rapidity of the air-current, diminishes 

 also the rapidity of cooling of the first disc. 



7. If the explanation I have just given is correct, it is evident 

 I ought to obtain a permanent sound by raising the wire-gauze 

 disc to a red heat by means of a galvanic current. The first 

 experiments undertaken to verify this result were unsuccessful. 

 I had not sufficiently considered the cooling effect of the moving 

 air on the wire-gauze, and the degree of intensity which the 

 galvanic current should have, in order that its calorific effect 

 might compensate for this refrigeration. It was only by making 

 use of thirty of Grove's cells combined so as to constitute a bat- 

 tery of six cells, that I had the satisfaction of obtaining a per- 

 manent sound. It was so loud that it could be easily heard 

 two or three rooms from the laboratory where the experiment 

 was being performed. Nevertheless, the wire-gauze was not 

 raised all over to a red heat. 



The number of vibrations of the sound produced was deter- 



