316 



Dr. T. Carnelley on the 



This Table shows that if the elements be taken in the order 

 of their atomic weights there are, with but very few excep- 

 tions, no sudden jumps from a high to a low melting-point, 

 or vice versa, whilst the end of one series of elements runs 

 gradually into the beginning of the next. Thus we pass by a 

 series of gradations from easily fusible sodium to almost infu- 

 sible silicon, and thence on to gaseous chlorine, then to easily 

 fusible potassium, &c. This running of the end of one series 

 into the beginning of the next is especially seen in the case of 

 group VIII. ; and it is worthy of note that a similar thing 

 occurs with the heat of formation of the dichlorides, and also 

 with the atomic magnetism of the same metals *, which latter 

 has recently been determined by Wiedemann (Phil. Mag. [5], 

 vol. iv. pp. 161, 276), thus : — 





Or. 



Mil. 



Fe. 



Co. 



Ni. 



Cu. 



Zb. 



Ga. 



Melting-point 

 Heat of for- 

 mation of 

 dicblorides. 

 Atomic mag- 

 netism 



a. 2270 



2170 



111990 

 100-4 



2080 



82050 

 83-1 



2070 



76480 

 67-2 



1870 



74536 

 30-5 



1364 



60988 

 10-8 



676 



(97200) 



303 



The above remarks will be quite sufficient to show that this 

 Periodic Law is a truly scientific one, and not a mere nume- 

 rical curiosity ; for it opens up new analogies, and therefore 

 points out new paths for the investigation of the elements. 



Numerical delations beticeen the Atomic Weights of the Ele- 

 ments. — Some interesting and curious numerical relations like- 

 wise exist between the atomic weights of the elements, of which 

 the following may be taken as examples. 



(1) The atomic weights of the elements of the first group 

 are simple multiples of 7*7, thus : — 







True 

 atomic 

 weight. 





Li ... 

 Na ... 

 K.... 

 Cu... 

 Kb ... 

 Ag.- 

 Cs ... 



? 



? 



Au... 



7-7 x 1= 7-7 

 7-7 X 3= 231 

 7-7 X 5= 38-5 

 7-7X 8= 616 

 7-7x11= 84-7 

 7-7x14=107-8 

 7-7x17 = 130-9 

 7-7x20 = 1540 

 7-7x23=177-1 

 7-7x26 = 200 2 



7-0 

 23-0 

 391 

 63-1 



85-2 

 107-7 

 1330 

 154-0? 

 181-0? 

 197-0 



[atom-analogues. 

 Calculated from the atomic weights of its 



Ought to be 199 according to Mendeljeff. 



Here the multipliers of 7*7, with the exception of those for 

 Li, Na, and K, form an arithmetical series with the common dif- 

 ference 3, whilst between Li and Na, Na and K the common 



* This was first pointed out by the author in a paper read before the 

 Royal Society, June 19th, 1879 (Proc. Roy. Sec. No. 197, 1879), 



