448 Messrs. Rowland and Hutchinson on the 



for the guard-plates, and 1 1 for the glass plates. L L L L are 

 collars for catching the oil from the bearings ; C C, C C/ are 

 speed-counters, C gear with the axle, and C C with C C in 

 the manner shown ; each has 200 teeth, and speed-reading is 

 taken every 40,000 revolutions. 



The needle system is enclosed in the brass tube T, ending 

 in the larger cylindrical box in which are the mirror and upper 

 needle. This is closed in by the conical mouth-piece Q, across 

 the opening of which is placed a wire grating. The mirror is 

 shown at M, the upper needle at N' and the lower at N. The 

 system is hung by a fibre-suspension about 30 centim. in 

 length, protected by a glass tube. The needle-system is 

 made by fitting two small square blocks of wood on an 

 aluminium wire ; on two sides of each of the wooden blocks 

 are cemented small scraps of highly magnetized watch-spring. 

 The needle thus made is about 1 x 1 x 10 millim. 



The mirror is fixed just below the upper needle, and is read 

 by a telescope 200 centim. distant. The plane of the mirror is 

 at an angle of 45° with the plane of the disks for convenience. 

 The whole is supported by the board 00 attached to a wall- 

 bracket. 



Two controlling magnets (WW) with their poles turned in 

 opposite directions are used. By means of the up and down 

 ■motion of either magnet, any change in the sensitiveness can 

 be attained ; and by the motion in azimuth, the zero point is 

 controlled. The advantage of its use lies in the extremely 

 delicate means it affords of changing the sensitiveness, much 

 more delicate than with a single maenet. 



The bed-plate B is screwed to one end of a table, at the 

 -other end of which a countershaft is placed (fig. 2). This is 

 run by a electric motor in the next room, the belt running 

 through the open doorway. The motor is 14 metres from the 

 needle. 



Although the disks and countershaft were carefully balanced 

 when first set up, and the table braced and weighted by a heavy 

 stone slab, yet at the speed used, 125 per second, the shaking 

 of the entire apparatus was considerable ; the needle was so 

 unsteady that it could not be read. This was seen to be due 

 to vibrations of the telescope itself and not to the needle. 

 To prevent it, each leg of the table on which the telescope 

 rested was set in a box about 30 centim. deep filled with 

 sawdust, and a heavy stone slab was placed on top of this 

 table. This entirely did away with the trouble ; the swing of 

 the needle was as regular when the apparatus was revolving 

 as when it was at rest. 



The two hard rubber rings (KB,) mentioned above have 



