194 ANNUAL RECORD OF SCIENCE AND INDUSTRY. 



the ear, and, therefore, that different co-vibrating parts of the 

 ear are set in action by the vibrations reaching the ear by 

 these two different routes, is a necessary consequence of my 

 hypothesis of the mode of audition, and was not suspected 

 until my hypothesis pointed it out to me; nor was it known 

 until I attempted to test the hypothesis by experiment. I 

 know of no other hypothesis which accounts for this fact, 

 which, while it is a necessary consequence of my own views, 

 is directly opposed to those hypotheses hitherto formed on 

 the mode of audition ; for, according to the latter, the co-vi- 

 brating parts of the ear make as many oscillations in a given 

 interval as the tympanic and basilar membranes." Amer. 

 Jour. JSci.j August, 1874. 



OX THE VIBRATION OF LIQUID SURFACES. 



M. Barthelemy has made a study of the undulations pro- 

 duced on liquid surfaces when these are set in vibration. His 

 best results were obtained when the vessel containing the 

 liquid was placed upon the resonant box of a vibrating tun- 

 ing-fork. In these circumstances the liquid surface is formed 

 of stationary elevations and depressions, the results of uni- 

 form vibration over its whole surface. The phenomena may 

 be observed either directly by a small telescope, or, better, 

 by projection upon a screen. For this purpose a beam of 

 sunlight is reflected obliquely from the liquid, and, passing 

 through a lens, the latter forms an image of the surface on a 

 screen. This image is so steady that it may be photographed. 

 In a rectangular vessel two sets of brilliant lines parallel to 

 its sides are produced, and bright points are formed at the 

 intersections of these lines. Barthelemy deduces the two fol- 

 lowing laws : 1. The breadth of the undulations is inversely 

 as the number of vibrations. 2. The distance between two 

 lines produced by the same fork is independent of the den- 

 sity of the liquid. In circular vessels the vibration figures 

 consist of equidistant concentric circles intersected by equi- 

 distant radii. If the vessel is vibrated by the fork touching 

 it, then a cross formed of quiescent liquid is produced : the 

 lines of the cross cut the vessel at its nodal points. Trian- 

 gular vessels give lines perpendicular to the sides, forming 

 brilliant hexagons, the centres of which are the angles of 

 fainter hexagons, having the radii of the first set for sides. As 



