REFRACTIVE IXDEX. 347 



solidify slowly, absorbed the liquid portion with filter paper, 

 extracted this with ether, and examined it in Abbe's refracto- 

 meter, an instrument which measures the angle of total reflection. 

 Skalweit examined this method and showed that it was important 

 to operate at a fixed temperature. 



Owing to the difficulty of maintaining a fixed temperature in 

 Abbe's refractometer, this method was not much used till special 

 instruments were devised. 



Amagat and Jean have devised an oleo-refractometer for deter- 

 mining the refractive index of oils and fats (Fig. 48) ; it consists 

 of a collimator, a hollow prism with sides inclined at an angle of 

 107°, and a telescope furnished with an arbitrary glass scale 

 placed in the focus of the eye-piece. In the collimator is placed 

 a piece of opaque substance, which cuts off the light from one- 

 half of the field. If the prism and the space outside between 

 the collimator and the telescope are filled with the same liquid, 

 there will be no refraction. If, however, the prism contains a 

 different liquid, the refraction will be indicated (in arbitrary 

 degrees) by the position of the junction between the li^'ht and 

 dark halves of the field on the scale. 



A standard oil (Indlc li/pe)* is supplied with the instrument, 

 and the scale is so adjusted as to read zero when this is placed 

 in the instrument. The oil or fat to be tested is placed in the 

 hollow prism and the position of the dividing line read off on 

 the scale. The temperature is kept constant l>y means of a 

 jacket, and is usually 45° C. 



Jean gives the following method for testing butter : — ^lelt 

 from 25 to 30 grammes of the buttei- in a porcelain dish at a 

 temperature not exceeding 50° C. ; stir well with a pinch or two 

 of gypsum, and allow to settle out at about the same tempera- 

 ture. Then decant the supernatant fat through a hot water 

 funnel plugged with cotton wool, and pour (while warm) into 

 the prism. Observe the deviation at 45°. Genuine butter gives 

 a deviation of about 30° to the left, while margarine gives 

 about 15°, and |coco-nut oil about 59" Lobry de Bruyn 

 has shown that genuine butters may show a deviation of 25° 

 to the left. 



It is evident that the addition of a mixture of coco-nut oil 

 and margarine would give a figure equal to that of butter. Muter 

 has, however, shown that the figure given in the oleo-refracto- 

 meter has a relation to the Eeichert figure, which would be much 

 disturbed by such a mixture. 



Muter's relation is expressed by Table LXXXI. 



* This is usually translated as " typical oil " ; the word " standard " is 

 more nearly equivalent to the French " type," than is " typical." 



