68 REPORTS ON INVESTIGATIONS AND PROJECTS. 



appear later. In all respects this instrument has proved thoroughly satis- 

 factory, and reflects much credit on the w^ork of our Pasadena instrument 

 shop. The large aperture of the collimating and camera lenses gives per- 

 fect illumination at the hmb of the sun, thus obviating a difficulty encoun- 

 tered in our previous work with the Rumford spectroheliograph. 



The driving mechanism, mercury flotation, methods of adjustment and 

 of setting on the spectral lines, and all other features of the design have 

 proved so satisfactory that no changes in the instrument have been 

 required. After a series of experiments, with from one to four prisms, we 

 finally decided to adopt two prisms for the daily work of the instrument, 

 since it was found that they would give admirable results with the lines 

 of calcium, hydrogen, and iron. The necessity of devoting this instru- 

 ment to routine work, and the heavy demands placed upon it by the 

 morning and afternoon program of observations, will prevent it from 

 being used for experimental purposes. However, this necessity was 

 foreseen ; in fact, it was hardly supposed that the 5-foot spectroheliograph 

 would serve so admirably for the routine work with hydrogen and iron 

 lines. 



A spectroheliograph of 8 inches aperture and 30 feet focal length was 

 designed for experimental work on limited regions of the solar surface, 

 with dark lines too narrow for use with the 5-foot spectroheliograph. The 

 construction of this instrument has been indefinitely delayed, through 

 the difficulty of making large prisms of sufficient homogeneity. Of three 

 prisms ordered from the Carl Zeiss Company in February, 1905, only one 

 has been delivered, and this proved entirely unsuitable. Indeed, it seems 

 possible that sufficiently homogeneous glass can not be obtained for the 

 purpose. For this reason an experiment was tried by Brashear in build- 

 ing up a prism of horizontal slabs of glass, carefully selected from tele- 

 scope disks of the same melting and free, because of their small thick- 

 ness (about one inch), from the difficulties of annealing encountered in 

 large prisms. The resulting prism was much more successful than the 

 single solid prism made by the Zeiss Company. Indeed, its resolving 

 power was sufficient for the purposes of the spectroheliograph, but the 

 amount of light scattered vertically, of no consequence in most classes 

 of investigation, was sufficient to prevent it from meeting the extreme 

 demands in this particular of the spectroheliographic work. It is ac- 

 cordingly being used, with excellent results, in our holographic studies. 

 I have now decided, in view of the difficulties encountered with solid 

 glass prisms, to try some experiments with fluid prisms, in the hope that 

 means will thus be found of completing the 30-foot spectroheliograph. 



My studies of the spectroheliograph plates, other than those made with 

 the globe-measuring machine (described in the work of the computing 



