418 
Analyses of Books . 
IJuiy, 
psychological character is made clear by even superficial observa- 
tion and comparison of the sheep and the tiger, or the eagle and 
the dove.” 
Yes, decidedly by “ superficial observation.” Whoever will 
closely study the habits of animals will find that, though vege- 
tarian species do not kill to eat, yet when possessed of sufficient 
strength and courage they frequently attack, and injure or kill, 
out of gratuitous malignity. It would be easy to name half-a- 
dozen species of Herbivora quite as dangerous as any beast of 
prey. 
We regret to find such serious flaws and oversights in Dr. 
Hamernik’s pamphlet, the more as in certain respedts he has our 
sympathies. 
Smithsonian Miscellaneous Collections. The Constants of Nature. 
Part V. A Re-calculation of the Atomic Weights. By F. 
Wigglesworth Clarke, S.B., Professor of Chemistry and 
Physics in the University of Cincinnati. Washington : 
Smithsonian Institution. 
Early this year it became our duty (“ Journal of Science,” 1882, 
p. 1) to discuss the generality of the principle of Evolution, and 
in particular the probability of its extension to the chemical ele- 
ments. In so doing, however, we considered it judicious not to 
lay any weight upon the law — or rather the hypothesis— of 
Prout, as not being sufficiently proven. The work before us 
tells decidedly in favour of that hypothesis, and as such is inte- 
resting not alone to the chemist and the physicist. 
It may be needful briefly to recapitulate Prout’s law. It as- 
sumes that, regarding the atomic weight of hydrogen = 1, the 
atomic weights of all the other elements will be multiples of 1 
by some whole number. The bearings of this hypothesis, if 
demonstrated, are admitted on all hands to be exceedingly mo- 
mentous. If the elements have existed independently from all 
eternity, if they have been formed — as the vulgar expression is — 
“ by chance, ” or if they have been created arbitrarily, the proba- 
bility of any definite relations existing between their atomic 
weights, or indeed any other of their properties, is vanishingly 
small. If, on the other hand, the elements have arisen under 
the adtion of some general law, if they mark out the successive 
stages at which permanence has been gained, we should then 
consider such relations as the one in question as a priori pro- 
bable. Hence it is but natural that Prout’s hypothesis should 
have been, from its promulgation in 1815 down to the present 
day, a disputed question. When it first appeared the determina- 
tions of the atomic weights, or, as they were then called, equiva- 
lents, were in the majority of cases of doubtful accuracy. Hence 
