522 Mr. A. Griffiths on the Source of 



Although the results of the second method correspond to 

 what should be expected from the first, they should probably 

 not be considered to prove more than that the ordinary laws 

 of reflexion hold at the surfaces of such thin films. 



Conclusions. — The conclusions to be drawn from the work, 

 briefly stated, are: — 



(1) That the thickness of the black film of a soap-solution 

 is not constant, and may vary from 6 /a/j, to 40 fifi. 



(2) That the film of a pure oleate solution may consist of 

 two black films, the thickness of the second being about half 

 the limiting thickness of the first, which is about 12 /z/x. 



(3) That the addition of glycerine or potassium nitrate to 

 a pure oleate solution prevents the appearance of the second 

 black film. While in the first case the thickness is greater, 

 in the second the range of the thickness is smaller. 



The observations by the first method have been repeated by 

 Mr. Chamberlain; the mean of his results for the thickness of 

 the second black film was 6'2 /x/x, and for the limiting thick- 

 ness of the first black film 11*2 /jl/jl. 



It is here desired to acknowledge indebtedness for many 

 ideas in both methods to Prof. Michelson ; and also to thank 

 him for his interest and encouragement in the work, as well 

 as for giving us an instrument so beautifully adapted to work 

 of this character. Many suggestions should also be acknow- 

 ledged to Associate Prof. Stratton, particularly in regard to 

 the apparatus used. Finally, it is desired to thank Dr. Millikan 

 for suggestions in the preparation of the manuscript. 



Ryerson Physical Laboratory, 

 University of Chicago, March 1st, 1899. 



XLIX. Note on the Source of Energy in Diffusive Convection. 

 By Albert Griffiths, M.Sc. (Vic), A.R.C.S. (Lond.)* 



AT the conclusion of a paper on " Diffusive Convection " t 

 the author, partly in the hope of producing a discussion, 

 asked certain questions relating indirectly to the source of 

 energy in the apparatus under consideration. 



After the publication of the paper in the Philosophical 

 Magazine, Prof. FitzGerald made some remarks on it in 

 i Nature,' and gave a concise account of the actions at work. 

 He pointed out, what was already known to the author, that 

 there is a tendenc}^ towards cooling when diffusion causes the 

 rise of the centre of gravity. Stimulated by Prof. Fitz Gerald's 



* Communicated by the Physical Society : read Feb. 24, 1899. 

 f Phil. Mag. s. 5. vol. xlvi. p. 453 (1898). 



