416 



SCIENCE. 



[N. S. Vol. XVI. No. 402. 



planned research must soon come to be a 

 recognized and a most important duty of 

 such schools that I made the laboratory for 

 research a leading feature of the scheme. 

 Gradually its scope and its work have ex- 

 tended until it has ultimately become one 

 of the most fruitful of all the adjuncts of 

 such institutions. To-day, in every engi- 

 neering school of importance, it is con- 

 sidered no less desirable and necessary to 

 provide for systematic research than for 

 laboratory instruction of students. Pro- 

 gress in the industries is now very greatly 

 promoted by the work of this department 

 of the professional school. 



VII. 



Study of the data and results of the in- 

 vestigation, thus collated and placed in a 

 form suitable for convenient and accvirate 

 examination and comparison, once the 

 work is complete so far as experiment is 

 concerned, is the final step of the investiga- 

 tion proper. If the object has been the 

 identification of an important datum, as 

 when our old friend, Rowland, measured 

 the relation of mechanical and thermal 

 energy, or as when my former pupil, 

 Michelson, measured the velocity of light, 

 the main work is that of correction of 

 minor errors and of standardization of ap- 

 paratus. 'V\'Tien the purpose is the deter- 

 mination of laws as well as facts, as when 

 my former colleague, Langiey, studied the 

 resistances of the air as a problem in avia- 

 tion, the niunerical values must not only be 

 checked and corrected, but the numerical 

 relations of those facts and the law repre- 

 sented must be formulated. The former 

 task involves skill in mechanical construc- 

 tion, the latter talent in the production of 

 a scientific theory; both, when Avell per- 

 formed, testify to genius in the investi- 



Formulation, tabulation and systematic 

 presentation for study thus may be so 



performed as to make the work of sur- 

 vey, of collaboration and of detection of re- 

 lations of law and of quantivalence com- 

 paratively easy. Lacking experience, or 

 talent, or system, the imperfect tabulation 

 of data, the inaccurate representation of 

 relations or the unintelligent grouping and 

 imperfect systematization of quantities by 

 the investigator may conceal rather than 

 reveal the solution of the problem in hand. 

 The best and principal protection against 

 such hindrance is correct and precise for- 

 mulation of the problem at the start. 

 Where a curve of results can be laid down, 

 or where any graphic or glyptic presenta- 

 tion can be made, it is usually easy to per- 

 ceive hitherto concealed relations and to 

 secure desired deductions and conclusions. 

 The production of lines and surfaces thus 

 exhibiting these smoothly continuous varia- 

 tions of value also has an exceedingly im- 

 portant use in the detection of individual 

 errors and the establishing of correct fig- 

 ures. 



Rumford, Davy, Mayer, Joule and Row- 

 land, thus establishing the measure of the 

 'mechanical equivalent of heat,' made reve- 

 lation of a fundamental datum on which, 

 coupled with the principle of the quantiv- 

 alence of the energies, it became possible 

 to biiild up a new science which should give 

 prophecy both of paths of progress and of 

 limits of improvement and of efficiency for 

 the heat-engines which have proved of in- 

 valuable service to the engineer and, 

 through him, to the world. Carnot, seer 

 and prophet, formulated in outline the new 

 science. Rankine, studying the facts and 

 data and fundamental principle thus re- 

 vealed, from the standpoint of the engi- 

 neer ; Clausius, as a mathematical physicist 

 malring a similar study, the single fact and 

 the single datum, combined with a single 

 law, that of the equivalence of thermal and 

 dynamic energy, enabled them, independ- 

 ently, each to accurately construct the new 



