abbot] new ccelostat and horizontal telescope 87 



fleeted beam is fixed, and about this point all other parts of the field 

 revolve, either with uniform or (as in the Foucault form) with a 

 velocity variable according to the position of the heavenly body. To 

 counterbalance these advantages the coelostat does not in its simplest 

 form send the beam in a fixed direction independent of the declina- 

 tion, but when used, as is customary, to give a horizontal beam, this 

 beam deviates toward the north of an east and west line for objects 

 south of the celestial equator and vice versa. Furthermore, when 

 sending a horizontal beam nearly eastward horizontally, it is clear 

 that the mirror is employed at a very unfavorable angle for objects 

 near the western horizon. 



It had been proposed here to get over these two disadvantages by 

 the use of a second plane mirror, itself mounted so as to permit of 

 moving it upon a U-shaped track with north and south branches close 

 to and on the east and west respectively of the polar axis carrying 

 the first or rotating mirror, which latter was intended to cast its 

 beam nearly horizontally to the east in the morning and toward the 

 west in the afternoon. This device was tried, but there were serious 

 objections to it, the chief one being that for objects far south and 

 near the meridian (like the sun at noon in December) the cross-sec- 

 tion of the reflected beam was very small compared with the aper- 

 ture of the first mirror. 



IMPROVED FORM OF TWO-MIRROR CCELOSTAT 



Fortunately a better device for solar work at this latitude was 

 then thought of. This consists in reflecting the beam due south from 

 the rotating mirror and thence due north from the second mirror 

 over the top of the first. The beam from the first mirror shoots 

 upward at an angle with the vertical equal to the sum of the angles 

 of latitude and declination ; and for the sun at Washington this angle 

 is about 62° at summer solstice and 16° at winter solstice. There- 

 fore to give a horizontal northerly directed beam the second mirror 

 is to be inclined forward 14° at the former period and 37° at the 

 latter.^ In the following table is given the diameter of mirrors for 

 this form of coelostat and for other instruments to furnish a fixed 

 horizontal north and south solar beam at the latitude of Washing- 



^ If it was necessary to incline the second mirror forward still more, a 

 special support would be required ; but it seems unquestionable that the 

 Ritchey support system {Astro physical Journal, w, 143, 1897), modified only 

 so far as that the mirror is stuck by cement or by exhaustion of the air to the 

 numerous balanced supporting pieces, would be quite as efficient here as for 

 the usual case of a mirror reclining face upward. 



