234 Mr. Michael Donovan on certain Improvements in 



from the material. They are short it is true ; but when the bars of the needle 

 are very close to the three acting faces of the coil, these fibres are sometimes 

 sufficiently long, resisting, and numerous, to obstruct the motions of the com- 

 pound needle when in a state of great sensibility. The following plan succeeded 

 in removing this source of uncertainty. 



The wire, well covered with silk, is wound on a brass frame, each round 

 so tight and close to the adjoining one, that the first layer may lie perfectly flat, 

 without springing or bellying in any part. When the first layer has been wound 

 from the beginning to the end of the frame, it is to be wound back again ; and 

 if the rounds of the first layer have been put on very compactly against each 

 other, they will support the second layer without allowing the wire to force 

 itself between any two which lie underneath. The flatness of the first layer 

 will insure the same perfection for the second ; and equal attention to tightness 

 and flatness in winding the third layer, or fourth, if there .be one, will produce 

 a coil of great regularity and closeness, and of equal thickness, — qualities of the 

 greatest utility. 



Under these circumstances, it is obvious that the frame must be strong 

 enough to sustain the collective tension of such a number of coils ; and there- 

 fore brass, well covered in the touching parts with hard varnish, is to be preferred 

 to ivory; although the latter is often used on account of its being a non-conductor. 

 The four sides, which constitute the parallelogram on which the coil is wound, 

 would not, without being inconveniently clumsy, afford sufficient resistance to 

 this tension, but that the two end pieces of the frame, instead of being mere 

 bars like the side pieces, constitute portions of a circle at the inner side, and 

 are flat at the outer. 



From what has been described, it is evident that the vacancy between the 

 layers of the coil being but one-eighth of an inch, the numerous filaments pro- 

 jecting from the silk or cotton must be even more likely than in ordinary cases 

 to obstruct the needle: and the very means adopted for securing sensibility 

 would also react against that result, but for the following construction. Two 

 plates of very thin, well-hammered brass are soldered to the two arches in such 

 a manner as to connect them, and constitute what may be called a floor and a 

 ceiling to the narrow chamber in which the lower bar of the compound needle 

 rotates, thus efiectually removing all possibility of obstruction by filaments, and 



