TKMPEEATURE OF THE SURFACE OF THE MOON. 33 



ing magnet sliding on a brass rod, and a system of five upper and five lower magnets connected 

 by an alununum rod with an alnminnin vane, the time of a single vibration without damping 

 magnet being G.5S seconds. 



In ineparation for the extremely delic^ate final work on the moon, the following changes were 

 made: (I have to express my great obligations to the kindness of Prof. Sir William Thomson 

 and of Professor Rowland for valuable suggestions.) The most important of these improvements 

 has been the replacing of the short fiber by one 33 centimeters in length (for the brass rod being 

 substituted a hollow glass one, in the center of which is the fiber) ; and, second, the reconstruction 

 of the needle. 



In the new astatic system constructed at this observatory in November, 1884, the aluminum 

 rod carrying the magnets was replaced by a hollow glass fiber. The aluminum vane, it occurred 

 to me to replace by au insect's wing, and one was most advantageously made of dragon-fly's wings, 

 (in which nature has supplied an admirably rigid and light construction). A minute platinum 

 paddle at the bottom of the glass fiber, touching the surface of oil in an oil-cup, was supplied, and 

 a new system of magnets. These are made by rolling soft sheet-steel, 0.076 millimeters thick and 

 5 millimeters wide and from 7 to 9.5 millimeters long, around a short straight piece of wire into 

 minute cylinders, carbonizing them in fused ferrocyanide of potassium, and tempering them in 

 mercury. The strength of one of the little magnets was found to be 874 Gaussian units, and of 

 these there are in all twelve, six on each system.* 



In forming the connections, it will be found advantageous to employ a battery of a consider- 

 able number of cells (e. g. twelve, of a gravity battery), and to reduce the current by interposing 

 resistance. Under these circumstances, it might appear that there was no advantage in using the 

 current from twelve cells over that of one, if the current were as strong in either case. Such, 

 however, is not the fact; for the accidental fluctuations due to the minute casual changes which 

 take place in the most constant cell are obviously equalized by the use of a current which is the 

 mean of that from a considerable number of cells.* Pains are taken to wrap every connection and 

 binding post in cotton, and a great number of minute precautious, which are not here detailed, 

 have been observed. 



The damping magnet is arranged so as to take any position between the bottom of the glass 

 rod and a point 1.46 meters above it, a graduated vertical scale being provided above the galva 

 nometer rod. The mirror of the instrument is a minute silver-ouglass concave reflector, of 1-meter 

 radius of curvature. The transparent scale, which is on the west, at 1 meter distance, is a portion 

 of a cylinder of 1- meter radius, and is graduated in millimeters from to 500. Accordingly, when 

 the needle points north and south, and the optical axis of the mirror east and west, the image is 

 at 250, at the middle of the scale. This image is a circle of light about 3 centimeters in diameter, 

 with a central vertical line (the shadow of a wire).t With these values, to carry the image wholly 

 off the scale demands a rotation of the needle through onlj' about 7 degrees. As a rule, this small 

 maximum deviation, with the employment of a curved scale, renders reduction for arc unnecessary 

 in such observations as these. The needle, when renderetl as astatic as possible, performs a single 

 vibration in about a minute; but in this condition the directive force is apt to vary from one day 

 to another, and the time of vibration, as a rule, to grow more ra^iid until a shorter period is reached^ 

 at which it becomes relatively constant. For the purpose of forming an approximate estimate of 

 the sensitiveness of the instrument, it may be stated that when making a single vibi-ation in 10 

 seconds a deflection of one millimeter division on the scale is given by a current approximately 

 equal to 0.0000000013 ampere. 



East of the galvanometer, and nearly in the prolongation ot the optical axis of the upper mir- 

 ror, are two small bar magnets, on an independent stand, a minute movement of either of which 

 serves to bring the image on to any point of the scale when necessary without altering the resist- 

 ance in the resistance box. 



* The device of the hollow magnets is cine to Mr. F. W. Very, of this Observatory, at whose suggestion also the 

 number of battery cells was increased with great advantage. The actual construction and astaticising of the needles 

 also has been chiefly due to Mr. Very's patience and skill. 



tThe employment of a telescope and a flat mirror, reflecting the inverted scale, is in some respects preferable to 

 this arrangement, which is continued in use, however, at present from its greater facility of adjustment. 

 S. Mis. 69 5 



