604 



SCIENCE. 



[N. S. Vol. XI. No. 277. 



sumptions whicla have afterwards proved 

 to be improbable or untenable, have brought 

 discredit on views, which in their essentials, 

 were of great value. Perhaps no hypothesis 

 has suffered more in this respect than the 

 atomic theory. In his book on the ' Con- 

 cepts of Modern Physics' (p. 85), Stallo 

 mentions certain points on which, he says, 

 all atomists are agreed. Among others 

 are these : " Atoms are absolutely simple, im- 

 changeable, indestructible; they are physically, 

 if not mathematically, indivisible.' ' 



Without speaking for the physicists, I 

 can assert most positively that none of these 

 attributes are in the least essential to the 

 conception of the chemical atom. Whether 

 the atoms be simple or complex, divisible or 

 indivisible, we have at present no satisfac- 

 tory means of deciding, and whether they 

 be one or the other, it in no wise affects the 

 conception of the atom as the chemical unit. 

 It is believed by some, on spectroscopic 

 evidence, that atoms are decomposed at the 

 high temperatures existing in certain stars, 

 and a similar explanation has been offered 

 for certain electrical phenomena exhibited 

 by gases. All we can say at present is, 

 that bj' no chemical or physical process 

 known to us, do atoms undergo division or 

 transformation to an extent appreciable by 

 chemical methods. An atom of carbon al- 

 ways acts with the combining weight 12; 

 if it consist of several independent parts, 

 we do not know it, because in all reactions 

 thus far known, these parts always act to- 

 gether. The idea of the transmutation of 

 the elements, while resting at present on a 

 very slender basis, is entirely justifiable as 

 a working hypothesis. 



The supposed indestructibility of the atom 

 amounts merely to this, that with our 

 limited range of experimental methods, we 

 have not been able to cause any appreciable 

 portion of matter to disappear as such per- 

 manently, but can always recover it un- 

 changed in mass and chemical properties. 



To assert that matter cannot, under any 

 circumstances, be made to disappear as 

 matter, seems to me to be the most unjusti- 

 fiable dogma imaginable. 



As to the unchangeableness of the atoms, 

 we are equally in the dark. That an 

 atom of oxygen in water is similar in every 

 respect to one in iron rust, we cannot assert. 

 There are certain physical properties of the 

 elements which persist in their compounds, 

 among which are the optical properties. 

 Every transparent compound has a defi- 

 nite molecular refractive power, and it has 

 been found that the figure representing this 

 may be divided up in such a way as to as- 

 sign a definite portion to each atom in the 

 molecule. If we once know the atomic re- 

 fractive powers of the several elements, we 

 can calculate with considerable accuracy 

 the molecular refraction of any compound 

 containing these elements. For example, 

 the molecular refraction of alcohol, C„HgO, 

 is the sum of twice the atomic refraction of 

 carbon, six times that of hydrogen and 

 once that of oxygen. JTow remarkably 

 there are certain exceptions. For instance, 

 oxygen combines in either of two ways, 

 which are called single and double union, 

 which are symbolically represented in the 

 case of carbon compounds, by C = O and 

 C — O. Its atomic refraction is different 

 in these two forms, and we cannot posi- 

 tively state that the difference is not due to 

 a temporary change of some kind in the 

 atom itself. At the same time (with the 

 above qualifications and some others of a 

 similar nature) the persistence of the 

 atomic refraction and certain other plij'sical 

 properties through all the combinations of 

 an atom, affords some indication that its 

 nature remains essentially unchanged. 



Finally, questions as to the shape, size, 

 hardness or penetrability of the atoms, are 

 matters of indifference to the chemist at 

 present. For his purposes they may be re- 

 garded simply as centers through which 



