221 
that they expanded with heat and contracted in cooling'. 
Had there been any exceptions to the rule, they must have 
been detected, for an elongation of one fifty-thousandth of an 
inch was easily read off ; and a tube of zinc, 4^ inches in 
length, could be observed from a temperature of 0° to 100° 
centigrade, without removing it from the apparatus ; and a 
change of 1 0° of temperature produced a change of more than 
26 divisions of a scale read oft' by means of a micrometer. 
In the course of a series of experiments made, in conjunc- 
tion with Professor F. Grace Calvert, upon a number of 
other metals and alloys, and detailed in a Paper recently read 
before the Royal Society, it was found that both zinc and 
steel have a very dift’erent ratio of expansion according as 
they are more or less crystalline in their texture. 
Thus, a bar of zinc 2| inches long, cast horizontally, 
expanded for 100° centigrade 216-5. 
The same zinc cast vertically, having two axes of crystalli- 
zation, 257*1 ; while the forged zinc tube expanded only 
150-0. 
The differences in steel are not so remarkable. 
The same length of steel at a maximum softness, 62-5. 
The same bar at a maximum hardness, 84-0. 
The same bar tempered so as to be malleable, 6 1 -4. 
The steel rods used in a pendulum tested by Mr. Joseph 
Baxendell, in the Observatory at Crumpsall, expanded 64. 
The zinc tube 150. 
It will be seen by calculation that 29*11 inches of zinc 
will expand suflBciently to compensate for the expansion of 
the steel rods, but an addition has to be made on account of 
a change which takes place in the elasticity of the pendulum 
spring. It becomes weaker in heat and stronger in cold, and 
this has the same effect as a little increase or decrease in the 
length of the pendulum. No rule has, I believe, been laid 
down as to the exact amount of additional expansion of the 
zinc tube to compensate for this ; but in former experiments 
