NATURAL PHILOSOPHY. 185 



ing its position and its distance from the centre of "Ti'iivity of the 

 cross section au<>"nientino;. As to the second law, the deviations 

 are not i^reater than may be attributed to difl'erences in the 

 moduli of elasticity of ditferent sticks, and the greater shifting 

 of the neutral axis of the sticks most strained. The third law 

 cannot be regarded as even approximately true, except in case 

 of sticks whose length bears a high proportion to their depth. 

 The fourth law fails as well as the third. Professor Norton has 

 made with the same apparatus a series of experiments on the 

 degree of set or residual deflection communicated to sticks by 

 varied strains and under varied circumstances, and obtained 

 interesting and curious if not important results. But these he 

 did not enter upon at this time. 



ELEMENTS OF MATTER. 



Prof. Charles A. Seely, at the meeting of the American Asso- 

 ciation, at Salem, in his paper on the Classification of the Elements 

 of Matter, after aUuding to some points in the philosophy of 

 classification in general, remarked that the revised atomic nota- 

 tion and the doctrine of atomicity are the foundation of what is 

 termed " modern chemistr}'." They are the starting-point of the 

 most reasonable explanations of facts. They test the genuine- 

 ness of the old grouping of the elements, and they lead us to 

 altogether new systems of grouping. Atomicity divides the 

 elements into two grand divisions, perissads (whose atoms com- 

 bine singly with an odd number of other atoms), and astiads 

 (whose atoms combine with an even number of other atoms). 

 The author supposed himself the first to observe that if the 

 classes be placed side by side, the members being arranged in 

 the order of their atomic numbers, and at the same time so that 

 those of corresponding atomic numbers be brought into juxtapo- 

 sition, then the elements of corresponding atomic numbers will 

 be found to be remarkably allied in phvsical properties ; in other 

 words, the astlad is paired with the perissad of corresponding 

 atomic number. It will also be observed that each member of 

 a natural group will be found opposite a member of a related 

 natural group. It thus appears that the physical properties of 

 the elements are closely related to and possibly dependent upon 

 the atomic weight and the atomicity ; the relation is so close that 

 the atomic weight and atomicity may be taken as the sole data 

 for a subdivision of the two grand classes. For the best realiza- 

 tion of this thought the author acknowledged indebtedness to 

 Prof. Charles S. Peirce, of Cambridge, and exhibited a diagram 

 of classification prepared by him. In this chart the elements are 

 represented at heights corresponding to their atomic numbers, 

 and the natural series appear in vertical columns ; in other 

 words, the ordinates of the points fixing the position of the ele- 

 ments represent the atomic weights, and the abscissas of the 

 members of a series are equal. The chart plainly exhibits the 

 fact of the pairing Ix'iween individuals and between well-recog- 



16* 



