TUB MICRO-SPECTROSCOPE. 129 



cases, the apparent magnitude of the effect of an increase 

 in thickness is far greater when the object is thin than 

 when thick, and past a certain thickness the change is 

 comparatively very slight. If only small crystals can bs 

 obtained, it is well to mount a number of different thick- 

 nesses ; but when it is possible to obtain crystals of suf- 

 ficient size, it is far better to make them into wedge- 

 shaped objects, since then the effect of gradual change in 

 thickness can easily be observed. Different kinds of 

 crystals require different treatment, but, as a general rule, 

 I find that it is best to grind them on moderately soft 

 Water-of-Ayr stone with a small quantity of water, which 

 soon becomes a saturated solution, and then to polish 

 them with a little rouge spread on paper laid over a flat 

 surface ; or else, in some cases, to dissolve off a thin layer 

 by carefully rubbing the crystal on moist blotting-paper 

 until the scratches are removed. Then, whenever it is 

 admissible, I mount the crystal on a glass, and also cover 

 it with a piece of thin glass with Canada balsam. Strongly 

 coloured solutions may be examined in test-tubes, or may 

 be kept sealed up in small bottles made out of glass tubes, 

 the light then examined being that which passes through 

 the centre of the tube from side to side. (Most of these 

 solutions require the addition of a little gum Arabic to 

 make them keep.) Such tubes may be laid on the ordinary 

 stage, or laid on the stage attached to the eye-piece. 

 Smaller quantities may be examined in cells cut out of 

 thick glass tubes, one side being fixed on the ordinary 

 glass with Canada balsam, like a microscopic object, and 

 the other covered with thin glass, which readily holds on 

 by capillary attraction, or may be cemented fast with gold 

 size or Canada balsam, if it be desirable to keep it as a 

 permanent object. Such tubes may be made of any length 

 that may be required for very slightly-coloured solutions. 

 Cells made out of spirit thermometer tubes, so as to be 

 about ^th of an inch in diameter, and an i ncn l n o> 

 are very suitable for the examination of very small quan- 

 tities ; but where plenty of material can be obtained, it 

 IB far better to use cells cut out of strong tubes, having 

 an interior diameter of about |-ths of an inch, cut wedge- 

 ehape, so that the thickness of the solution may be JtJi 



