Primitive Crystals of Quartz, and the Sulphate of Barytes. 239 



is the finding of crystals that are adapted to it. It is requisite not 

 only that their planes should be brilliant, but also that the image of 

 the line or bar reflected upon them should be perfectly defined ; its 

 edges should be seen with an exactness which they who have not 

 given attention to the subject will scarcely believe to exist on sur- 

 faces so small. No reliance can be placed upon a solitary ad- 

 measurement taken upon the natural planes of one crystal, however 

 brilliant and perfect they may appear ; it is essential that several 

 should be found to exhibit a perfect agreement, before it can be 

 assumed that the actual value of any angle has been obtained. 

 From one crystal of quartz, nearly perfect at both terminations and 

 affording pretty clear reflections, I obtained six admeasurements of 

 the same angle from its various planes. One of 94° 9', another 

 of 94° 12', a third of 94° 16', and three of 94° IT. I did not 

 discover one crystal whose planes were sufficiently perfect to enable 

 me to obtain corresponding measurements from plane to plane, 

 around it. 



It cannot however, be doubted, that much of the labour attend- 

 ant on the use of the reflecting goniometer in the measuring of the 

 angles of crystals on their natural planes, will be spared, and that 

 the results will be much more satisfactory, because more accurate, 

 whenever cleavages can be obtained sufficiently brilliant for its use. 

 In almost every instance in which I have tried it upon the planes 

 produced by good fractures, the success has been complete. The 

 great value of this instrument, which may be used with readiness 

 and ease, demands the attention of every one who has the slightest 

 pretension to crystallographical research. I feel much pleasure in 

 adding an instance of the accuracy of which it is susceptible. Hav- 

 ing seen in some periodical publication, a notice that Dr. Wollaston 

 had discovered differences in the value of the angles of the primi- 



2 h 2 



