276 Messrs. P. G. Parker and A. J. Dalladay on a 



a minimum, — we should expect two more such points where 

 the solution is supersaturated with respect to its water- 

 content. 



Hence (5) our equation of state must represent these 

 points. (6) At the cryohydrie point all these four points 

 must coalesce into one ; this also must be deducible from 

 our equation of state. Similar arguments apply to the 

 complementary osmotic pressure P 6 . 



In putting out these notes, I have had the invaluable 

 assistance of Mr. G. W. Walker, F.P.S., to whom my best 

 thanks are due. 



XXVI. On a Precision Method of uniting Optical Glass — 

 the Union of Glass in Optical Contact by Heat Treatment. 

 By R. G-. Parker, B.Sc, A.C.G.L, and A. J. Dalladay*. 



[Plate V.] 



OWING to the numerous defects of the cements available 

 for constructing glass cells for the examination of the 

 optical properties of liquids, the following process, by which 

 a much more perfect joint between two pieces of glass can 

 be obtained, has been developed. Briefly, the surfaces of 

 each of two pieces of glass are so worked and polished that 

 they are either both plane or of the same curvature to a very 

 high degree of accuracy. They can then be placed together 

 in " optical contact," and if they are subjected to carefully 

 controlled heating will unite and become one piece. 



In order that this method may be of any value, it is 

 essential that the temperature at which union takes place 

 shall not only be far below the melting-point of the glass, 

 but also below the annealing temperature — for even at this 

 temperature very slight strain would cause sufficient defor- 

 mation of worked glass surfaces to render them useless. 



It is very difficult to place the windows of a cemented cell 

 sufficiently accurately parallel for certain work depending on 

 the phenomena of the interference of light ; parallelism to 

 + 1 (one) minute is only attained with considerable trouble, 

 especially if the adhesive material requires baking to cause 

 it to harden completely. The wedge shape of the enclosed 

 liquid may be too great for interference work, even if the 

 angle between the windows is not greater than a few 

 seconds. 



* Read before the Faraday Society^on December 18, 1916. 



