4 Wilde, Atomic Weights of Neon, Argon, Krypton, Xenon. 



natural varieties of organic species determined by 

 naturalists." 



It is not a little remarkable that, while the laws of 

 chemical combination and of crystallisation are established 

 on strictly mathematical bases, the multiple proportions 

 of the elements among themselves have not, so far, been 

 placed on a similar foundation, for the reason, principally, 

 that some of the experimental atomic weights are not 

 whole numbers. 



Now the constancy of the angles in crystals of the 

 same substance is considered to be a fundamental law of 

 crystallisation. I have found, however, by exact measure- 

 ments with the goniometer, that small differences exist 

 in the angles of apparently perfect cubical crystals of 

 fluorite (amounting in some specimens to more than one 

 degree), analogous to those observed among the atomic 

 weights. Similar small differences may also be observed 

 in the angles of other forms of crystals. Such residual 

 interferences, arising from known and unknown causes, 

 would rightly, not be admitted by mineralogists as invali- 

 dating the general laws of crystallisation. 



Addendum. 



It is a common error to assume that discoverers in 

 various departments of science are, necessarily, authorities 

 on the co-ordination of the subject of their discoveries 

 with the general properties of bodies, and with the real 

 nature of things. Thus (i) Berzelius assigned to silicium 

 an atomic weight equal to 21, while the accepted 

 number is 28. Regnault has, however, pointed out that 

 the atomic weight of this element should be 35, in order 

 that it may enter into the law of the atomic heat of other 

 elements. (2) Peligot adopted 1 20 as the atomic weight 



