390 MAGNETIC OBSERVATORY AT 



time is measured, and from which the ordinates of the curve are sub- 

 sequently read off. The cover is next put on, and the lens-frame put 

 in position by stops on the base, to which the supports of cylinder 

 and the clock are attached. Observing the place where the point of 

 light now strikes the paper, a pencil mark is made, then the time 

 noted, and a corresponding reading taken in the telescope on the 

 fixed scale, which for that purpose is illuminated by a candle held in 

 the hand. 



After the lapse of 24 hours, less about 5 minutes, a similar reading 

 is taken, the time noted, and a corresponding pencil mark made on 

 the trace. The cylinder is then lifted out, the sheet taken off, and 

 a new one immediately put on and started as before. 



The former sheet is then floated on a saturated solution of gallic 

 acid, to which a small quantity of the nitrate of silver solution has 

 been added, and allowed to remain from 10 to 20 minutes, until the 

 trace is distinctly brought out. After being next rinsed several times 

 with pure water, it is soaked for 15 minutes in a solution of liyposul- 

 pltite of soda — one ounce to ten ounces of water — to fix the trace; 

 and, lastly, the excess of hyposulphite is thoroughly washed off, and 

 the sheet allowed to remain in pure water until next day's operations 

 have brought the following sheet to the same stage. The former 

 sheet is then allowed to drain, and is thoroughly dried, exposed to 

 the air on a flat piece of glass, to which it adheres, and is thus kept 

 from wrinkling or warping. 



TABULATING THE RECORD. 



The trace is usually a Avell defined black line of about the twentieth 

 part of an inch in breadth, with somewhat pale margins or edges. 

 The middle of the line can readily be estimated to the nearest hun- 

 dredth of an inch by a practised reader. Having the time and scale 

 readings corresponding to the beginning and end of the trace, and 

 the direction of the time-scale, it is easy to lay off points at each end 

 of the sheet corresponding to a certain average scale-reading — 600— 

 through which points a line of abscissa? is drawn with pencil. In 

 order to read off the ordinates or scale-reading for every half hour, or 

 oftener, when the character of the trace makes it desirable, a scale 

 of half hours is applied to line of abscissa?, and the corresponding 

 ordinates read off on the edge of a square divided into spaces corres- 

 ponding to minutes of arc for the declinometer, and to one ten-thou- 

 sandth part of the force for the other two instruments. 



The process of preparing the paper and subsequent operations are, 

 of course, precisely the same for the three instruments. All three 

 sheets are taken off and replaced in rapid succession before the bring- 

 ing out of the traces is proceeded with. 



BIFILAR MAGNETOMETER OR HORIZONTAL FORCE INSTRUMENT. 



This instrument is designed to exhibit and record the variations in 

 the horizontal component of the earth's magnetic force. To that end 

 the magnet is constrained by a twisted suspension by two threads to 



