From Horace Darwin to G. H. Darwin 14 April [1875]

Dear George

I did not express myself properly in my letter, there would be no difficulty about getting the steel rod consentric with brass cylinder if it was large enough to turn it conveniently but I was trying to get it only sufficiently thick to be strong enough & then it would be difficult to turn, the reason I was trying to get it as thin as possible was on account of friction, wh. I think will be a difficulty, especially as you think the cylinder will tend to slew & then friction will be indeterminate; but now the machine is shorter it will be easier. I will designe the machine therefore on the old designe only with the steel rod plenty thick enough; O will talk to the foreman turner, (a cute man) about what thickness will do, & then the machine will be like your present designe in principle only there will be less friction; I will consult the foreman what he thinks would be the best material, as I think it might be made of all wrought iron, all steel, all gun metal, or steel rod cast in gun metal, or steel rod fastened into gun metal by screw or several other ways.

I should suggest that you try the following experiment. Fix your machine up & fill it with pitch, wh. I should have thought might have been done by putting lumps of pitch into it & putting it on the Hob for a bit & then take the double brass [dome] through wh. the steel rod passes off so that the brass [DIAG HERE] cylinyer is quite freel & floating in the pitch as it were. Then tie a bit of string to the top of the cylinder & carry it up some way & fasten it to a support virtically over the cylinder & strain is so that it will about take the weight of the cylinder, & put a swivle in it, wh. on second thought will not be necessary I think; this will allow the cylinder to turn absolutely with out friction if only you can keep the cylinder in its place, I think this might do it; anyhow I think it wd. be worth trying, as if by any chance it does act it will make your results so much more trustworthy   My picture represents a view of the machine looking down wards, steel rod black in the middle; there are four strings coming from the [DIAG HERE] brass wheel A1B1 A2B2 . . . . ; these strings pass over 4 pulley A1 .. & to them are attached 4 equal weights   this would tend to stop it moving sideways & slightly any tilting too, but if rod a b (in former picture) picture was long & string b, c short, I think it would also tend to keep it straight, in fact you migh have no string & only a swivle. I’m inclined to think this would act, I’m pleased with it.

You must make out if 4 strings from brass wheel are the best no. any no more then two would do I suppose. I also suggest that you rute a white line in a radial direction on the surface of the pitch when you try your experiments to see into what sort of a curve it turns

H Dar