These files of correspondence contain letters from Sir Francis Galton to Frederick Smith, and chiefly relate to work the two men were carrying out regarding a consistent method of measuring horses from photographs. During this period, Smith was working at the Woolwich Remount Depot, and was preparing a paper titled ‘Relation between the Weight of a Horse and its Weight-carrying Power’. It seems that Smith intended to have this paper presented at the Royal Society, but it was eventually published in the Journal of Comparative Pathology and Therapeutics in 1899. [A reprint of this article can be found at FS/2/3/1/3]. Galton was preparing an article titled ‘Photographic Measurement of Horses and other animals’, published in ‘Nature’ in 1898, and a copy of this article is catalogued at FS/3/3/2/6.

There are no letters from Smith to Galton found in this collection. Galton’s notepaper all has a black border, indicating a recent death, and this is assumed to be for Galton’s wife, Louisa Jane Butler, who died in 1897.

1 – Letter to Frederick Smith from Francis Galton, 27 Nov 1897

/0 Comments/in /by
Terms of Use
The copyright of this material belongs to descendants of the creator. Images are permitted for reuse under a Creative Commons, Attribution, Non-commercial license.

[FS/3/3/2/1]
[[1]]

42 Rutland Gate Rd

Nov 27/97

Dear Captain Smith,

Since writing & talking with you about horses, I have not being [sic] idle but have been puzzling in various directions, and now want to consult you once again. Let me first say that I have had a good work at the Records of American Trotting horses, and finished a memoir thereon, not yet printed. It seems that there exists a very valuable collection of facts there, that could be used for hereditary inquiry, but which is at present practically inaccessible,

[[2]]

though with a little management it might to all appearance be easily obtained. I am in full correspondence about this, & only mention it now to shew that I am hopefully busy.

What I want now to ask you about, relates to the former question of horse – measurement, and to the following idea.

I suppose there to be little difficulty in gradually initialising a custom of having a “common” photograph taken of each valuable horse in the way to be described[.] I mean quite independently of the artistic and costly photos now usual.

[[3]]

The “common” photo is intended to carry on its face the necessary data for making measurements in the photograph itself, and these seem to be very sensible.

  1. for the body, a stiff leather & dark plate, with a white (or blue, which is white photographically & less staring?) rectangle in paper of say 20 inches x 10 inches attached to it, which might be graduated & bear a recognised standard mark[.]
  2. for the head, a graduated headstall (lengthways.

[[4]]

It seems to me from looking at horses in the street in harness, & those with waterproof rugs over their backs, that a plate of the sort I mean would practically coincide with a vertical plane tangential to the horse’s side. *postscript [struck-through text illeg] there would be no [illeg] difficulty in getting correct measures however much (within reasonable limits) the vertical position of the horse was oblique to the line of slight[.] For, by prolonging in pencil in the photograph of the top and bottom lines, seen in perspective, a vertical scale suitable to that perspective is obtained, and the horizontal scale can be easily deduced. Thus all the [vertic] distances in a vertical plane tangential to the side of the horse can be measured from the photo.

[[5]]

The graduated head stall would give a similar means of measuring the head; whatever its obliqueness to this line of sight might be[.]

3/ In measuring the body, the points in it might be spotted, with a small stiff brush dipped in chalk and water, before photographing. Precautions being of course insisted on, that these should be done by a skilled & trustworthy person[.]

It seems to my very enthusiastic

[[6]]

mind, that the interest & utility of providing means for future accurate measurement of every valuable horse, would lead in time (if the movement were first stabled in respect to the winners at prize shows) to its pretty general adoption[.] The points to be dotted with chalk and water are those we talked about, subject of course to re-consideration.

How does all this strike you? I should truly be obliged to free criticism, not wishing to make any move in a subject on which I have so little

[[7]]

practical knowledge, without approval from those who have.

Very faithfully yours with kind remembrances to Mrs Smith

Francis Galton

To avoid prolixity I have perhaps not entered fully enough into the ways of getting a scale from the graduations in the plate. The plate may slope vertically as well as in the way described but a divergence of even 8⁰ makes only percent error!

The data for finding all out, are the measured lengths in the photograph of AB, AC, and CD (page 4).

[[8]]

The scale for any measurement in the tan vertical tangential plane to the horses side can be deduced from these data, either by graphic methods such as architectural draughtsmen use, or perhaps more simply by the help of tables for the purpose. I worked out many of these during another & different attempt with horse – photographers, but have not regularly thought these out those that are suitable for the present purpose. There seems no difficulty in doing so.

, , , ,

This work is in:

open all | close all

2 – Letter to Frederick Smith from Francis Galton, 1 Dec 1897

/0 Comments/in /by
Terms of Use
The copyright of this material belongs to descendants of the creator. Images are permitted for reuse under a Creative Commons, Attribution, Non-commercial license.

[FS/3/3/2/2]
[[1]]

42 Rutland Gate, S.W.

Dear Captain Smith

You are most obliging & helpful. Let me suggest in addition (if not too late) a white streak down the line of the shoulder & possibly elsewhere.

I am making a little, optical instrument that promises to act very well, to standardise colour. Could you manage conveniently

[[2]]

to get me a few snips of hair, ches:t bay, brown & black, and of 2 or 3 shades of each, for me to mount experimentally[.] As much of the whole lot as could be stuffed into a common envelope, would suffice.

Very faithfully yours

Francis Galton

, , ,

This work is in:

open all | close all

3 – Letter to Frederick Smith from Francis Galton, 7 Dec 1897

/0 Comments/in /by
Terms of Use
The copyright of this material belongs to descendants of the creator. Images are permitted for reuse under a Creative Commons, Attribution, Non-commercial license.

[FS/3/3/2/3]

42 Rutland Gate, S.W.

Dec 7/97

Dear Captain Smith

Last night on returning from a few days in the country, I found your most welcome photographs; the card about A reached me this morning.

The measurements are most easy to make on the photo and it is just possible that they may be

[[2]]

more trustworthy when taken as the perspective projection

[sketch of horse and camera positions]

of the spots on a vertical plane touching the horses side; that at is as mn (not MN). What is wanted is an horizontal projection of those

[sketch of camera position]
[[3]]

points viz μ, v. The small correction necessary to find these is easily to be obtained from the data h, k, Mμ, Nv.

h & k might be recorded by the photographer (a rough approximation sufficing) and Mμ Nv might be guessed for the particular horse after having been measured for two or three types of horses[.] The correction wd be very small never I should think exceeding 1 inch, except in the height of withers & elsewhere along the spine of the back[.]

the values of mn & your

[[4]]

measures of MN are given below

[table of data including mn (from photo) and MN (from measures)]

I took photos (not yet developed) of 2 horses in the country but foolishly did not measure between the spots, my object being to practice the technique of manipulation. It was an opportunity lost.

When I have worked at these, I should like to consult you as

[[5]]

to the best way of getting numerous photos with measures of horses; and of determining what points to select for measures.

Your gummed discs are beautifully clear. I am inclined to lithograph & cut out discs on this principle,

[sketch of disc]

(I have no compasses or scale at hand, & have used a shilling to make the circle. It shd have been 1” in diameter.). or else to stencil them. Pray accept my heartiest thanks.

[[6]]

As regards the whistles I enclose one for you to keep. You will probably find it to work best when the piston is greased.

There is so much difficulty in determining for each individual whistle, the relation between the depth & the note, that I would suggest your discarding the question altogether & in its place to find the depth that cats can just hear & some few children also. If in a more intelligible measure than

[[7]]

no of vibrations. Shaw’s method of doing this was by the method of interference between the sound wave, as it travelled towards a screen, and its echo as it travelled back. By moving the jet of a sensitive flame to and fro, the nodes of quiescence were easily found & the distance between 2 adjacent nodes was quickly found, whence the no of vibrations per sec: was of course found at once.

In using whistles for high notes, recollect that any intervening obstacle casts a well defined sound shadow.

[[8]]

The sound does not travel round them freely, as it does with ordinary notes. I tried all the beasts in the zoo with a whistle at the end of a walking stick. An india rubber ball was squeezed in the hand, and it

[sketch of whistle]

expelled air that passed through a thin tube to the whistle, which gave a tiny squeak. The big carnivores hated it. Some human persons do so also. It gives to them a strange dizzy sensation, that is not sound, but is conveyed by the ear in a different form to the brain.

Very faithfully yours

Francis Galton

P.S.
Do not call me “Dr.”

, , , , ,

This work is in:

open all | close all