78 JOURNAL OF THE 



law and is in accord with our fullest krxOwled<4-e. The 

 second part of the law as usually stated, that these 

 properties are periodic functions, attempts in a measure 

 to define the dependence. It may also be true but it is 

 not fully proved and is open to objections. It seems to 

 mo that this hypothetical portion could well be left in 

 abeyance until fuller knowledg-e g'ave it a stron<»-er foot- 

 in<>', meanwhile substituting- something- less open to crit- 

 icism and which cannot weaken the central truth. 



Take this table and examine it. First we lind two 

 kinds of periods made use of — periods containing- seven 

 elements and those containing seventeen. If it had 

 only been possible to arrang-e all of the elements in 

 sevens as Newlands attempted to do, the periodic idea 

 would have been most convincing- and the law of oct- 

 aves running- throug-h nature would have seemed most 

 wonderful. But these elements do not admit of being- 

 arrang-ed in this way and the use of periods of differ- 

 ent leng'ths is to fresh young- minds, unacquainted with 

 mathematical expedients, somewhat forced. 



Secondly, there is a very anomalous position assig-ned 

 to the triads or, as sometimes written, the tetrads, Fe, 

 Co, Ni, (Cu,) etc. They have been set off to them- 

 selves, clearly so as to make the other elements fall 

 even approximately into their places and into the proper 

 sevens. I say approximately, for the student soon sees 

 that althoug-h there is a similarit}" there is also a wide 

 difference between the elements of the lirst and of the 

 last seven in any x>eriod of seventeen. 



Thirdly, in the lower periods, in order to g-et ele- 

 ments to fall into their places a g-reat many unknown 

 elements have to be interposed. Thus between cerium 

 and ytterbium, the next element in the list, there are 

 blank places for sixteen elements. The third large 



