ADVISORY COMMITTEE ON ASTRONOMY 1 29 



The extensive equipment of American observatories has been 

 alluded to in the report. Without important additions of apparatus,, 

 twice or three times the present number of workers could find profit- 

 able employment for many years in making observ^ations with ex- 

 isting instruments, and reducing them. But this fact by no means 

 proves that our present equipment, powerful though it be, is not 

 open to radical improvements, in which may lie the greatest hope 

 of future progress. The history of science has showm that a single 

 improvement in instruments may render years of work with inferior 

 apparatus unnecessary. This is particularly true of astrophysical 

 research, which, in view of the nature of the phenomena investi- 

 gated and the possibility of employing the innumerable devices of 

 the physicist, cannot be considered as subject to precisely the same 

 conditions that obtain in researches dealing exclusively with the 

 positions and motions of the heavenly bodies. In such researches 

 the instruments and observational methods are comparatively few 

 in number, and in the course of years a definite system of procedure, 

 departed from only in matters of detail, has been gradually evolved. 

 But astrophysics, which is so closely akin to physics, may be re- 

 garded almost in the aspect of an experimental science, susceptible 

 of indefinite growth in method and in scope. No sharp line can be 

 drawn separating its investigations from those of astronomy on the 

 one hand or those of physics on the other. The primary interest of 

 the astrophysicist may be an astronomical one ; but it may equally 

 well be that of the physicist, utilizing stellar temperatures and 

 pressures to solve problems of atomic structure for which laboratory 

 appliances are inadequate. 



Under these circumstances it will be understood that as long as 

 astrophysics continues to be a growing science the provision of new 

 instruments for its researches will be necessary. A concrete exam- 

 ple, drawn from an investigation of the evolution of the red stars, 

 will illustrate the nature of the instrumental needs that constantly 

 arise in the present state of the subject. 



The study of stellar evolution is based upon the spectroscopic 

 analysis of starlight. With our largest telescopes it is perfectly, 

 possible to photograph the spectra of stars much fainter than those 

 at the limit of naked-eye vision. Measurements of the positions of 

 the lines in these spectra determine the chemical composition of the 

 star's atmosphere, set an upper limit to the pressure in it, and give 

 an accurate value of the star's velocity in the direction of the earth. 

 In most cases such spectra can be arranged in a definite series. 



