127 
by frost was obviated, as the experiments were performed 
upon a solid granite rock in situ, and this could not be 
hardened by cold to any considerable extent so as to affect 
the results, and yet in these experiments the rails are shown 
“ to exhibit only from one- third to one-fourth the strength 
at 10° Fahr. which they possessed at 84° Fahr.” 
Dr. Joule observed that the admitted fact that the sup- 
ports of the bars in Mr. Sandberg’s experiments were in a 
different condition at the two temperatures rendered the 
results arrived at with them valueless as evidence on the 
question at issue. 
“Further Observations on the Strength of Garden Nails,” 
by J. P. Joule, D.C.L., F.RS., &c. 
Since communicating the paper on the Alleged Influence 
of Cold in giving Brittleness to Iron, I have collated the 
results with cast iron nails in order to show the range of 
strength in such specimens. 
Height of Fall 
Percentage of 
of Hammer. 
Fractures. 
2 inches 
0 
24 
,, .......... 
0 
3 
JJ 
6-25 
34 
v 
23-5 
4 
JJ 
30 
44 

36-4 
54 
?? 
37-5 
64 
?? 
48 
7 
62'5 
7A 
' 2 
}> 
64-3 
84 
J? 
75 
10 

92-8 
I chose the garden nails for experiment after some thought, 
as presenting a marked variety of metal in contrast with the 
iron and steel wire, tempered and untempered. I did not 
expect them to possess great strength, but having found 
them to require a heavier blow than I expected to fracture 
them, I have had the curiosity to make some experiments 
on them which may be interesting to the Society. 
I took pairs of the nails, placed them head to point 
parallel to each other so that pressure applied in the middle 
by pincers sufficiently forcibly would fracture one of them. 
