GLEANINGS 
IN 
SCIENCE. 
No. 36 . — December, 1831 . 
I . — Expansion of Metals by Heat. By J. Prinsep, Esq. r. R. s. 
Having been requested by Lieutenant Wilcox, tlie officer lately in chnrge of a 
trigonometrical survey of the Brahmaputra river, to ascertain the rate of expansion 
of the brass rods used in measuring the base of his series of triangles at Rangamati, 
I think it will not be uninteresting to put on record a brief account of my experi- 
ments, although they may not offer any great novelty either in the process adopted* 
or in the results obtained, and although the near accordance of the latter with 
former experiments of a like nature, can hardly hope to be regarded as any con- 
firmation of the labours of the eminent philosophers who have, at different times, 
devoted their attention to similar researches, with means of much superior 
accuracy. 
Lieutenant Wilcox's measuring rods are formed of brass wire, one quarter of an 
inch in thickness, and twenty-five feet in length, supported by a trussed frame-work 
of deal-wood, constructed on the plan adopted by Captain Herbert in his Hima- 
layan survey ; one of them is represented in plate XVIII. fig. 1 : t lie curve given to 
the longitudinal pieces of the wooden frame is the same as they were found to 
assume when laid horizontally, supported only at the two extremities. The 
scantling of the wood- work, one inch square, is not sufficient to give perfect stiffness 
to the frame, even when supported by a tripod in the centre as represented in the 
plate, but the brass rod is itself dressed in a horizontal position, by means of a fine 
wire extended along the axis nearly in contact with it ; the weight employed to 
stretch this wire is seen hanging down to the right of the frame. 
To measure the dilatation of so long a rod, it was necessary to provide the means 
°f maintaining its whole length of an equable temperature, at two distant and 
determinate points of the thermometric scale. The experiments of Roy and 
Troughton, and those of Lavoisier and Laplace were made on bars of five and six 
feet in length respectively. These were immersed in ice for the lower fixed point, 
a nd in a long trough of water, kept at the boiling point by means of a number o 
lamps, for the upper degree of the scale. This apparatus was found inconvenient 
* n management, especially at the boiling point, which it was difficult to maintain 
^ith regularity : it w T as evidently quite inapplicable for a bar of fi' e tiim s t e e ng , 
t° say nothing of the expence of constructing its various parts, t ere ^ 
cht) ice of the pervading action of steam as a means of communicating the Life 
te »iperature ; and for the lower, I was constrained to be content wi a s ream 
of Water at 82°. The apparatus fitted up for the experiment!! ait lr '' l0n ° 
th «e two agents is represented in fig. 2. A B is a leaden pipe of one 
N EVV SERIES, NO. IX. 
