Ill 
transverse strain either of weight or pressure, and it appears 
to me that if this point can be satisfactorily settled it will 
go a long way in settling the other points now in dispute. 
As my object, in showing that I have in these experi- 
ments eliminated as far as seems possible all sources of error, 
will be best effected by minute detail, you will excuse any- 
thing that may seem trifling. As I was not trying the 
absolute strength of any sort of cast iron I did not see the 
force of Mr. Brockbank’s objection to my using Jin. bars 
instead of the orthodox lin. bars. I could obtain Jin. bars 
equally good castings, and the machinery for breaking them 
was more manageable and in my opinion more exact. 
Messrs. Rye, Son, and Ogden, of Newton Heath, kindly 
undertook to make for me 50 bars, each 3ft. long by Jin. 
square, all out of one ladle, and of No. 3 Glengarnock pig 
and Kirkless Hall common pig — I name these although it 
does not seem of importance ; all I wanted was good, sound, 
clean, and equal castings; and, knowing the purpose for 
which they were intended, with great care they turned 
them out so good that not one of those sent to me was 
rejected. I now cut each of these bars into tlmee lengths 
of lfo. each, and as they were cut they were thrown into a 
heap making nearly 150 pieces. They were now taken and 
all their ends covered with paint, in order that the new 
fracture might be examined as they were broken. The heap 
was then brought into the laboratory, having thus had three 
chances of perfect mixing. A boy of 11 years of age now 
handed me the pieces singly from the heap, and as I 
received them I placed them alternately one by one in two 
lots, until I had got 70 pieces in each lot. One of these was 
now taken and put into a cask capable of holding 2cwt. to 
3cwt. of freezing mixture composed of pounded ice and 
chloride of sodium (which instantly reduces the temperatui'e 
to zero), and being surrounded with sawdust, they were 
kept there for nearly 4S hours. 
