216 APPENDIX. 



a phenomenon complementary to the above. The circle 

 now presents four patches of coloured light, one in each 

 quadrant ; and we generally see near the centre four 

 black or obscure spots, which correspond to the arms of 

 the cross in the other position. 



Such is an outline of the microscopic appearances 

 presented by these little crystals, which are probably the 

 minutest bodies in which so complicated an optical struc- 

 ture has hitherto been witnessed. I find that the smaller 

 circles are, the more perfect is their form, and the 

 brighter their colours. 



These crystals, as I have already observed, probably 

 consist of spicula diverging from a point, but which are 

 in the closest possible contact, and in a state of complete 

 mechanical cohesion. It seems to follow as a conse- 

 quence from such a structure, that their density must 

 increase from their circumference towards their centre. 

 Now it is worthy of remark, that Sir David Brewster has 

 discovered very similar phenomena by polarized light in 

 the crystalline lenses of certain fishes, which are known 

 by direct experiment to increase in density towards the 

 centre. Indeed the figure which he has given of the 

 lens of the codfish in the Philosophical Transactions for 

 1816 (Plate XII. fig. 1,) is so like the appearance of one 

 of the crystals which I have described, that it might be 

 supposed to have been intended for a representation of it. 



Having pointed out this resemblance, I may also men- 

 tion another class of facts to which I think those I have 

 described possess a considerable analogy. I mean the 



