INVERTEBRATA, CRYPTOGAMIA, MICROSCOPY, ETC. 149 



which are kept apart to any desired distance by indiarubber. The 

 plates — say 3" X I4", and about ^V inch thick — shoukl be of hard 

 brass, so as to bear screwing up without bending, and so causing leakage. 

 The curve of the trough may vary, but the semicircular — say 1^" x f " 

 — is the most economical, as the two plates can be turned in the lathe 

 at once, and this shape is also a convenient one for use. The bevel 

 should be wide, so as to allow objectives to work close to the curve. 

 The screws should be of equal hmgth, that the trough may be level 



\ \ 



and steady when reversed, and be long enough to allow the plates to 

 be set about y\ inch apart. The nuts should be milled, and as small 

 as can be conveniently handled. The glasses may be semicircular, 

 but oblong ones 2" X 1" seem to answer equally well. It might be an 

 advantage to cement the lower one to the plate, but this is not essen- 

 tial. The edge of the bottom glass should be even with the edge of 

 the plate, and the upper one about ^ inch from edge of plate, which 

 will be found convenient when filling troughs from a dij)ping tube, &c. 



The best thing to separate the glasses is half of a circular flat 

 indiarubber band of the thickness which will give the required 

 distance between the glasses. These bands, notwithstanding the 

 sulphur, &c., contained in them, have not been found to act injuriously 

 on living organisms. 



The trough may be put together, and then used in the usnal way, 

 or, as is generally better, the following plan may be adopted : — The 

 lower glass and rubber band being in their place on the lower plate, 

 the object is arranged on the glass with needles in a little water, so 

 that it may be best examined, and the upper glass and plate being put 

 on, the whole should be screwed np, care being taken that the pressure 

 (which need not be excessive) is equal, and the plates kept parallel ; 

 the trough can then be filled up with water. 



Teasdale's Test Slide for Dark-Ground Illumination. — This con- 

 sists of ten or more parallel lines, 2000 to the inch, disposed as a 

 symmetrical pattern of eccentric radials. The lines are ruled imder 

 perfect equality of circumstance, and with a conchoidal fracture and 

 not a " cut," so as to give a maximum of brilliancy. 



On applying this test, if the field be equally illuminated by the 

 spot lens or ccntrally-stopixnl condenser, the bands will appear all 

 equally bright; but if the illumination be in any way faulty or one- 

 sided the eye instantly detects it, the ])ands being so ruled that those 

 which lie at right angles to incident unilateral light are best seen. 



