TKANSACTIONS OF SECTION A. 535 



for scrapin"- tlie colouring matter off the copper rollers in calico printing. Tho. 

 piece used°was 75 centimetres long, and was cut to a breadth of about 2 centi- 

 metres. A brass weight of about 200 grammes was securely soldered to one end 

 of it, and the spring was bent like the spring of a hanging bell, to such a shape 

 that when held firmly by one end the spring stood out approximately ma 

 straic'ht line, havmg the weight at the other end. If the spring had no weight the 

 curvature, when free from stress, must be in simple proportion to the distance along 

 the curve from the end at which the weight is attached, in order that when 

 held by one end it may be straightened by the weight fixed at the other end. _ 



The weight is about 2 per cent, heavier than that which would keep the sprmg 

 straight when horizontal ; and the fixed end of it is so held that the spring stands, 

 not horizontal, but inclined at a slope of about 1 in 5, with the weighted end 

 above the level of the fixed end. In this position the equilibrium is very nearly 

 unstable. A definite sighted position has been chosen for the weight, relatively to 

 a mark rigidly connected to the fixed end of the spring, fulfillmg the condition that 

 in this position the equilibrium is stable at all the temperatures for which it has 

 hitherto been tested ; while the position of unstable equilibrium is only a few 

 millimetres above it for the highest temperature for which the instrument has been 

 tested, which is about 16° C. ■, -, j. o ^^ 



The fixed end is rigidly attached to one end of a brass tube, about » centi- 

 metres diameter, surrounding the spring and weight, and closed at the upper end 

 of the incline by a glass plate through which the weight is viewed. The tube is 

 fixed to the hypotenuse of a right-angled triangle of sheet brass, of which one leg, 

 inclined to it at an angle of about 1/5 radian, is approximately horizontal, and is 

 supported by a transverse trunnion resting on fixed V's under the lower end of the 

 tube, with a micrometer screw under the short, approximately vertical, leg of the 



triangle. » 



The observation consists in finding the number of turns and parts of a turn ot 

 the micrometer screw, required to bring the instrument from the position at which 

 the bubble of the spirit-level is between its proper marks, to the position which 

 equihbrates the spring-borne weight, with a mark upon it exactly in line with a 

 chosen divisional line on a little scale of 20 half-millimetres, fixed in this tube m the 

 vertical plane perpendicular to its length. 



The instrument is, as is to be expected, exceedingly sensitive to changes ot 

 temperature. An elevation of temperature of 1° 0. diminishes the Young's modulus 

 of the german silver so much that about a turn and a half of the micrometer screw 

 (lowering the upper end of the tube at the rate of 2/3 millimetre per turn) pro- 

 duces the requisite change of adjustment for the balanced position of the movable 

 weight. About 1| turn of the screw corresponds to a difference of 1/5000 m the 

 force of gravitv, and the sensibility of the instrument is amply valid for 1/40 of 

 this amount ; that is to say, for 1/200,000 diflerence in the force of gravity. Hence 

 it is not want of sensibility in the instrument that can prevent its measuring 

 differences of gravity to the 1/100,000 ; but to attain this degree of minuteness it 

 win be necessary to know the temperature of the sprmg to within 1/20° G. I do 

 not see that there can, be any great difiiculty in achieving the thermal adjustnaent 

 by the aid of a water jacket and a delicate thermometer. To facilitate the requisite 

 thermal adjustment I propose, in a new instrument of which 1 shall immediately 

 commence the construction, to substitute for the brass tube a long double girder of 

 copper (because of the high thermal conductivity of copper), by which sufficient 

 uniformity of temperature along the spring throughout the mainly effective pprtion 

 of its length, and up to near the sighted end, shall be secured. The water jacket 

 will secure a slight enough variation of temperature to allow the absolute tem- 

 perature to be indicated by the thermometer with, I believe, the required 

 accuracy. 



16. Comparison of the Harcourt and Methven Photometric Standards. 

 By W. Stepney Eawson, M.A. 



The author of the paper explained the growing necessity for a reliable govern- 



