Photography at Greenwich Observatory. 21 



curve is used. A separate gas flame is placed, so that its 

 light, reflected by a prism through a small cylindrical lens on 

 the top of the horizontal cylinder, forms another spot of light 

 on it, which remains stationary ; and as the paper travels this 

 prints a strong line round the cylinder, which, when the paper 

 is unrolled, becomes the base from which the distance of the 

 curve can be measured, and its value estimated, as next to be 

 described. 



In order to read off the indications thus obtained, and 

 translate the photographic curve into ordinary language, 

 observations are made with the theodolite in the old way, four 

 times a day, and these give the value of the indications of the 

 curve at those particular times, from which the value of other 

 distances of the trace from the base line can readily be mea- 

 sured by means of a scale drawn upon pasteboard. The length 

 of the paper being not always alike, and the going of the 

 clock likewise slightly irregular, it is necessary to have a 

 time scale as well, instead of simply dividing the paper into 

 hourly parts. This is effected by shutting off the light occa- 

 sionally for a few minutes from the cylinder, which, of course, 

 leaves a white spot in the curve, and the interval between two 

 such operations being accurately noted, gives the time equal 

 to a certain length of paper, and to divide this readily an 

 ingenious expedient is adopted. A slip of vulcanized India- 

 rubber is stretched in a brass frame, which, by a screw, will 

 lengthen or contract the slip, and the scale being marked on 

 the India-rubber, it can be altered so as to make its divisions 

 correspond with the value of the period of time measured by 

 two breaks in the curve on the paper. 



"With respect to the absolute length of the photographic 

 traces, it may be mentioned that a variation of 1° in the decli- 

 nation magnet is measured by five inches on the cylinder, and 

 that a variation of 1000th part of the horizontal force covers 

 about \ of an inch on the paper, and the same variation in the 

 vertical force about \ an inch. 



The same cylinder is made to record the indications of two 

 instruments, by being so placed that the light from each falls 

 on different sides of it, the base-line being made between 

 them. The declination magnetometer and horizontal-force 

 magnetometer record their variations on one horizontal cylin- 

 der, and the vertical- force magnetometer on a vertical cylinder, 

 which also receives the trace of the barometer obtained in the 

 following manner : — 



The instrument is a large bore one of the syphon form, 

 having the lower surface of the mercury more than one inch in 

 •diameter. This surface supports a glass float, from which rises 

 a vertical rod. This rod presses at right angles against a long 



