January 31, 19(i2.] 



SCIENCE. 



171 



generality and importance. First, what 

 laws connect the properties of compounds 

 with their composition 1 Secondly, what 

 laws govern the transformations of energy 

 during chemical changed Along each of 

 these lines a large amount of work has been 

 done, mostly empirical; and some regulari- 

 ties, some minor laws, are already recog- 

 nized. Systematically, however, neither 

 field is well known, and both offer rich 

 prizes to the investigator. Great masses of 

 more or less available data now exist; but 

 rarely do we find any group adequately de- 

 veloped. The determination of constants or 

 the measurement of thermochemieal rela- 

 tions is tedious in the extreme ; but a vast 

 amount of such work needs to be done 

 under some definite system or plan. At 

 present we have a datum here and a datum 

 there; some one in Germany makes a few 

 measurements, some one in Prance, or Eng- 

 land, or America makes a few more; but 

 seldom is there any attempt at cooperation, 

 and the isolated facts do not always fit to- 

 gether. The thermochemieal data are espe- 

 cially difficult to determine accurately, and 

 still more difficult to discuss in such a way 

 as to develop any clearly defined law. In- 

 deed, thermochemistry, of late years, has 

 fallen out of favor; for to many chemists, 

 despite its promise, it seems to lead no- 

 where. But laws must exist under all these 

 troubling questions, and we cannot despair 

 of their discovery. We can accomplish 

 little, however, unless we consider each of 

 the four great fundamental problems with 

 reference to the others, for they are sepa- 

 rable only in theory. Scientific research is 

 not linear, step following step in regular 

 succession; it is a network, rather, whose 

 interlacing threads are woven into patterns 

 of infinite variety. We trace individual 

 fibers, we see, more or less clearly, a part 

 of the design ; and this is the most that any 

 one of us can ever hope to do. 

 Now, whether we regard the fundamental 



questions of chemistry as four in number, 

 or condense them into two, we can use our 

 classification as an aid to research. Success 

 in the latter means a wise selection of prob- 

 lems, a choice which is conditioned by our 

 strength and our resources; but the first 

 step is to understand the bearings of what 

 we are trying to do. Whether our purposes 

 are modest or ambitious, our work must 

 have an influence upon that of others, and 

 the broader the plan upon which it is con- 

 ceived, the better the outcome will be. One 

 bullet well aimed is worth more than a 

 volley at random. One fact with a purpose 

 outweighs a hundred scattering observa- 

 tions. We may well ask, therefore, what 

 investigations are most needed by chemis- 

 try to-day? 



First, as to the nature of matter, with all 

 that that question implies. Taking all 

 kinds of matter into consideration, and 

 starting with the established distinction be- 

 tween elements and compounds, it would 

 seem to be obvious that work is most im- 

 peratively needed where our information is 

 least complete. Some elements, some classes 

 of compounds, have been much more ex- 

 haustively studied than others ; they, there- 

 fore, can best bear a temporary neglect, our 

 attention, in the meanwhile, being concen- 

 trated elsewhere. I do not mean by this 

 that any kind of research should cease, only 

 that each department should assume some- 

 thing like reasonable proportions. To or- 

 ganic chemistry, for example, we are in- 

 debted for many methods of research, and 

 for theoretical conceptions of great fer- 

 tility; but it is now time to apply them to 

 inorganic substances, and to see whether 

 they are generally valid. Whatever result 

 is reached, organic chemistry itself will be 

 the gainer; enriched by new suggestions 

 and resting upon firmer foundations, its 

 future advancement can be made all the 

 more certain. Meanwhile, carbon com- 

 pounds, by virtue of their serial relations, 



