490 Messrs. T. and A. Gray and J. J. Dobbie. [Apr: 24, 



is shown in the accompanying drawing, fig. 1, which is taken 

 from the previous paper. The glass which in each case was in 

 the form of a small globular flask of about 7 centims. diameter, 

 and from 2 to 3 millims. thick, was filled up to nearly the 

 bottom of the neck with mercury, and immersed to the same level in 

 mercury contained in an outer vessel. This outer vessel was 

 contained in a sand-bath, which could be heated by means of a 

 Bunsen burner placed below. The flask had in each case a neck 

 several inches long, and one extremity of a wire I was passed down 

 the neck so as to dip into the mercury within, while the other was 

 connected through a sensitive galvanometer with a battery of about 

 one hundred Daniell's cells. The other terminal of the battery was 

 connected with the mercury in the outer vessel. To enable the 

 current to be readily reversed, which was done between every 

 successive pair of readings, a reversing key (not shown in the 

 drawing) of somewhat novel construction, and of high insulation, 

 was placed in the galvanometer stand and interposed in the wires 

 joining the battery with the apparatus. It consisted simply of four 

 slender pillars of vulcanite attached to a base piece of the same 

 material. The top of each pillar was thicker than the stem, and had 

 a hollow scooped out of it to receive mercury. In one side of the cup 

 thus made, a piece of copper wire was fixed so as to terminate at one 

 end in the interior of the cup, and at the other to give a projecting 

 piece to which the battery wire could be attached. The hollows 

 were filled with mercury and cross-bridges of copper wire used to 

 connect the cups directly or diagonally, so as to give a current in one 

 direction or the other as required. 



The galvanometer used was the instrument described in the " Proc. 

 Roy. Soc." vol. 36, p. 287, to which we refer for a full descrip- 

 tion, but the following brief account of the instrument may be 

 conveniently given here. It consists of two pairs of coils with 

 hollow cores arranged so that the axes of each pair are parallel and 

 in a vertical plane. The coils act on a needle system of two horse- 

 shoe magnets of thin steel wire connected by a very light frame of 

 aluminium, and hung with their planes vertical, so that a horseshoe 

 corresponds to each pair of coils, and has its poles within the hollow 

 cores. Each pair of coils is in this instrument carried by a brass 

 plate, and these plates are set so as to make an angle with one 

 another of about 106°. The needles are not plane, but curved round 

 so as each to lie nearly in a cylindrical surface, the axis of which is 

 the suspension fibre, and which passes through the cores of the coils 

 without touching the coils on either side. Thus the needles can 

 move in the hollow cores through considerable displacements without 

 danger of coming into contact with the coils, and the cores are made 

 smaller than would be otherwise possible. The needles enter the 



