" Law " in Physical Optics. 

 Table II. (continued). 



329 



Substance. 



Temp. 



Density. 



{*»> 



/3= 



(«*-i>-(Pi-iK 



V 



at lower 

 temp. 





Allyl-acetyl-acetone . . . 



o 



60 

 99-8 



13-5 



99-2 



1-04248 

 100615 



0-97671 



0-89832 



1-46691 

 1-44386 



1-46156 



1-41868 



0292 

 0-291 

 0-309 

 0350 



0-96146 

 10238 



0-326 

 01709 



0314 



01515 

 01614 

 0-1680 

 0-1894 

 0-205 



0-175 

















Tables I. and II. contain the results yielded by substances 

 of very varied chemical types. It may be stated at once 

 that, although the last six substances have been put in the 

 Table, the discussion of the figures they yield will be deferred. 

 It is known that they are of very exceptional character ; for 

 example, that they change in chemical constitution when 

 heated, so that it is not strictly the same substance we are 

 dealing with at the two temperatures given in the Table*. 

 Excluding these compounds, the values of ft are of a satis- 

 factory order. They do not vary largely, the extreme values 

 being 0*09 and 043, most of them lying between 0*065 and 

 0*045. That is to say, on the hypothesis which led to the 

 calculation, the actual space filled by one gramme of the 

 substance is about one twentieth of a cubic centimetre. 



Column 7 shows the value of /3/v, that is the ratio of 

 the actual to the apparent volume of a liquid. The extremes 

 are 0' 15 and 0'054. If w r e exclude methylene di-iodide, a 

 substance of exceptional density, the extremes are 0'09 and 

 0*054, and twelve out of the sixteen liquids he between 0*076 

 and 0*054 ; the average value is 0*064. 



These results are promising, both with regard to the abso- 

 lute magnitude and the approach to similarity of value for 

 the different liquids. This last fact was to be expected, for it 

 is probable that ordinary liquids, under similar conditions of 

 temperature and pressure, have volume constitutions of a 

 fairly uniform type. 



It may be objected that the average values obtained for j3 

 and /3/v are much too small, seeing that other estimates of 

 * Compare Perkin, Chem. Soc. Journ. 1892, p. 840. 



