18

CORRESPONDENCE.

I have done my best to reduce some of the experiments; but the re-sults are so much at variance that I am completely puzzled. What addsgreatly to the difficulty, is their want of adaptation to mathematicalrequirements.

I shall not have it in my power to leave this place before the end ofnext week, without doing the parties a great injustice, but will do all Ican on my return. I have been very unwell since my return from Lon-don , in consequence of the cold I caught there.

I mentioned to you before, that there are fundamental experimentsnecessary to make any useful application of these of yours; but while Iam here I can do nothing, and before I should be at liberty to do any-thing, but try to reduce the experiments made, and suggest some modi-fications of them, it would be necessary to get the sanction of Mr.Stephenson, I suppose.

Were I at liberty to make any experiments, I would begin with somecylinders as free from rivets as possible near to the middle, so that theyshould break in a part not riveted. This might be done by taking longplates for the middle of the cylinder, and lengthening it out by smallplates, since the rivets in them would not signify. I would have threecylinders made, all of the same diameter and length, say 18 or 20 inchesdiameter at the least, and the length 16 or 17 times the diameter. Oneof these cylinders should have its plates half an inch thick, and theothers £th and ^th of an inch. I would have themsimple cylinders with one row of rivets on each side—none at the bottom or top near to the middle—norany aperture made there for the shackle. A sectionof the cylinder would be as here represented, whereI would rather there were no rivets; but as theycannot be avoided, they must be introduced. Thelength of the cylinder would be as below.

Fig. 11.

Fig. 12.

o v j'b o o o v v oT j o'fli j o oW

I would have three other cylinders made, of half the diameter andlength as the preceding one; and their plates should be of half thethickness as those above, say ^th, ^th, -Jj-th of an inch ; these cylindersshould be riveted on the side only, as in the former case.

If you think well to prepare these for me, I will make the experiments

4

4

&

4

A