l6o CARNEGIE INSTITUTION 



wholly lost. In this way, in an unsteady atmosphere, our photo- 

 graphic refractor of 32 centimeters aperture gives practicahy the 

 same results as the great refractor of 80 centimeters aperture. If a 

 greater ratio of aperture to focal length is chosen, say 1 : 10 — quite 

 apart from the great thickness of glass required in the refractor — 

 new difficulties result, in that the aperture of the collimator can be 

 only 1 : 10, and consequently for a given size of prism the collimator 

 will be too short. 



At a high station these several difficulties are not to be feared. 

 If you find a station with very quiet images, I should recommend 

 for spectrographic work — either with very high dispersion on the 

 brightest stars, c. g., for the determination of the solar parallax, or 

 with smaller dispersion on faint nebulae and stars — a reflector with 

 great aperture, and a ratio of aperture to focal length of about 1 : 30 

 or still less. I should also choose such a construction as to permit 

 the spectrograph to remain in a constant position, with the collimator 

 horizontal or in the direction of the earth's axis. Such a spectro- 

 graph with long collimator, very short camera, and very high dis- 

 persion would be best adapted to carry out the very important in- 

 vestigation on the motions within nebulae to which I recently called 

 attention. Our apparatus here did not permit me to accomplish 

 much in this direction. If the spectrograph were built in a fixed 

 position, not suspended from the eye end of the telescope tube, it 

 could be made much more stable and also easily maintained at a con- 

 stant temperature. It would thus be possible to employ very long 

 exposure times. A horizontal mounting is also to be recommended 

 for spectroheliographs of the largest dimensions. It is important to 

 make these photographs with a very large solar image, in order, for 

 example, to be able to study with precision the motions within a 

 sun spot. Spectrographic studies of the zodiacal light and the 

 aurora should also be included in the program. 



I heartily wish success to the great undertaking, and it will give 

 me pleasure if I can aid it in any possible way. 



[From E. Walter Maunder, Esq., Royal Observatory , Greenwich^ 



May 14, 1903. 



With regard to the question of astrophysical research, my own 

 position has led my thoughts in two directions. 



To me sun spots seem to be the most important subjects of stud}'. 

 Our work at Greenwich consists, as you well know, in taking two 



