342 Dr. J. H. Vincent on 



amount of Just on the refracting devices spoilt the experiment. 

 JDust on the surface of the water only spoils the appearance of 

 the photographs and does not matter much. 



The refraction was produced by introducing plates of glass 

 cut into appropriate shapes into the trough. The depth of the 

 water over these devices was generally about a fifth of a 

 millim. The depth was adjusted by trial until it was great 

 enough to allow of the propagation of the ripples. This 

 could be ascertained after a little practice by touching the 

 liquid surface with a clean metal pointer. If the liquid was 

 too shallow, it allowed itself to be heaped up and drawn about 

 like a viscous fluid. It is necessary to have the glass devices 

 quite clean, or else the surface of the liquid breaks, and the 

 whole arrangement has to be readjusted. The object in using 

 glass devices instead of metal ones was to enable the outline 

 of the shallow region to register itself on the photographs, and 

 also to get the waves delineated in the shallow region as well as 

 elsewhere. The cutting of the circular portions of the glass 

 devices was generally done by Mr. Pye on a machine which 

 he made for the purpose. This machine consists of a turn- 

 table upon which the glass plate is fixed, and an arrangement 

 for holding a diamond in contact with the plate while the 

 latter is rotated. 



Before commencing the actual experiments the capability 

 of the method was tested in a very simple way. A photo- 

 graphic plate was placed in a vessel with its film upwards. 

 Over this a sheet of glass was propped by three pieces of 

 metal made exactly the same height. Water was put into 

 the trough until it just flooded the plate. When the surface 

 was disturbed by ripples of a frequency about 60, these could 

 be photographed on the plate by instantaneous illumination 

 with parallel light. Refraction effects could bo recorded in 

 this way, though of course the picture was lacking in contrast. 

 Such pictures are not real photographs, they are merely per- 

 manent diagrams of the illumination at the bottom of the 

 trough, and not objective views of the surface. Many expe- 

 riments were carried out in this way, which affords a very 

 simple method of measurement of ripples. A noticeable cha- 

 racteristic of these pictui'es is that often some one ripple comes 

 out perfectly clear as a dark thin line on the negative. The 

 curvature of the ripple-crest happens to be just that required to 

 bring the parallel light to a linear focus by refraction at the 

 surface. 



The refraction method of taking ripple-photographs is much 

 less sensitive to small changes in the surface-shape than the 

 method of reflexion. Thus, suppose an element of the surface 



