306 PHOTOGRAPHY OF SOUND WAVES. 



In fig. 1(», PI. Ill, Avc hiivc tlu> refniction of the wiivc in the s:inu> tank 

 under oblique incidence. The bending of the wave within the tank 

 is very marked. The wave fronts reflected from the side which fol- 

 lows the unreflected portion is also interesting in connection with 

 Lloyd's single mirror interference experiment (No. 2 of series). 



After several failures 1 succeeded in constructing a prism with its 

 two refracting faces of this exceedingly thin collodion, which, when 

 tilled with cailtoiiic acid, showed the bending of the wave front exactly 

 as we tigui-e it in diagrams for light. It was necessary to have the 

 collodion thinner tiiaii before, since if we are to photogra])h the wave 

 after twice traversing the film, we must lose as little energy as pos- 

 sible by reflection. Fig. 11. JM. III. shows th(^ refraction in a carbonic- 

 acid prism, the biMiding luMng pai'ticularly noticeal)le in No. 4, on 

 wiiicii 1 lia\-e. with a i)air of dividers, ti-aced out the position which 

 the wa\'e front would liav(> occu])ied had it not trav»M"sed the prism. 



The IxMiding of the wave fi'oiit in the opposite direction is shown in 

 fig. li'. 1*1. III. wliei-e the same })rism is filled with hydrogen gas, in 

 which sound tiavels faster than in air. 



In tlie n«*xt flgure we have a very interesting case, though, owing to 

 the experimental ditKculties. the photographs are not quite as satis- 

 factory as somi^ of the others. It represents the transformation of a 

 spherical into a plane wave by pas.sage through a double con\'ex lens. 



The construction of the cylindrical lens of exceeding!}'- thin collodion 

 was a matter of great difticidty. The Hat. circular ends were made of 

 thin mica as free from stri.e as possible, that the passage of the wave 

 through the lens could be followed. On these disks the collodion flli« 

 was wound, the whole forming a hollow drum, which was then fllled 

 with carbonic acid. The soiuid wave, started at the principal focus of 

 this lens, is seen to be quite flat after its emergence (fig. 13, PI. III). 



^^'e will next take up some cases of difi'raction, beginning with the 

 well-known principle of Huygens, that any small portion of a wave 

 front can be considered as the center of a secondary disturbance and 

 that a small portion of this secondary disturbance can act as a new 

 center in its turn. 



In fig. l-t, PI. Ill, we have the wave starting above a plate with a 

 narrow slit in it. This slit is seen to be the center of a secondary 

 hemicylindrical wave which moves down precisely as if the spark were 

 located at the slit. After proceeding a short distance this secondary 

 wave encounters a second slit, and the same thing happens as before, 

 the little slice that gets through spreading out into a complete wave, 

 while the intercepted portion bounces back and forth between the 

 plates. 



Fig. 15, PL IV, shows the ver}' limited extent to which sound shad- 

 ows are formed. The wave is intercepted b}- a small glass plat(\ fJust 



