Imperial Institute of FroTice, 313 



Thclmation ravs previously polarized, we find two positions 

 dianietricallv opoosite, in which not a single luminous 

 molecule is 'reflected. Lastly, if we suppose thai, the cir- 

 cumstances benig the same, the g'ass nnrror is ligliled by rays 

 alreadv modified, by a convenient plate of rock cr\stal, tor 

 example, we shall see it successively tinged at each demi- 

 revolution with the whole series of prismatic colour';, as 

 well by transmission as by reflecticm— with this pariiculanly, 

 that at the same instant these two classes of colours will 

 be different. 



A ray of light which traverses two crystals, the princi- 

 pal sect'ions of which are parallel or perpendicular, furnishes 

 only two emersent rays. If a third crystal is interposed 

 betvveen the tv^o first, we see at least four images, txeept 

 in the case where the principal section ot this new crvsial 

 is perpendicular or parallel to the principal section of one 

 of the two others : hence results a very simple mclhod of 

 ascertaining when a substance, whatever may be its ex- 

 ternal form in other respects, has the double refraction : 

 lor this purpose, we have only to make it turn between two 

 rhomboids properly placed, and to observe if in certain 

 positions we see four images : novi' this is what hajipens, 

 for instance, with plates of mica, as M. Arago ascertauied ; 

 EO that this substance ought to be added to the list of iliose 

 in which mineralogists "have recognised the property of 

 doubling images. M. Arago thiriks however, that we can- 

 not rely entirely on the method in qtie-tion, in order to as- 

 certain when a body has the double refraction ; for, among 

 other consequences, it will result from his experiments, 

 that flmt glass ought to be admitted into the class of cry- 

 stallized bodies which double images, since certain plates 

 of this glass, interposed between tlie two rhomboids, act as 

 if they had an extraordinary plan of refraction, but of 

 which' the direction is different in different parts oi the 

 plate. Jt is perhaps right to remark, that M. Arago has 

 ascertained this property^iu pieces of flint glass more than 

 half an inch thick, and through which no double image 

 was seen, although the prismatic form was given to thfin ; 

 nay, he has someTimes, although rarely, met with tragmtnts 

 of the same glass, which as well as mica, su'phate ol lime, 

 &c. act variously upon rays of different colours; but these 

 properlits seem to ari-;e from some peculiar circumstances 

 in the fusion of flint gla^B, for the fragments which possess 

 them are not numerous. 



When we interpose a plate of mica of a certain thick- 

 f\ts&, between two rhomboids the principal sections of 



wluch 



