January 6, 1898] 



NATURE 



2X1 



graphs of horses and other pedigree stock could be rendered 

 available for strict scientific studies in heredity, the material 

 is copious, and as it would in time extend through many 

 generations, should far exceed in value anything that is 

 now procurable for those purposes. But all depends on 

 that " if." The basis of science is exact measurement, for 

 which the existing photographs are unsuitable. My present 

 object is to show that by paying strict regard to conditions of 

 a simple kind, an ordinary photograph will be transformed 

 from a mere picture into a record of real scientific value ; so 

 that if photographs should hereafter be habitually made of 

 pedigree stock (not only of horses) under those conditions, and 

 be afterwards published, a mass of material would quickly ac- 

 cumulate, sufficient to advance the science of breeding far be- 

 yond the point at which it now stands. Artistic photographs 



vn 



"6 CtMn^rtc 



Fig. I. 



are not to be discouraged. Their object is to exhibit animals 

 in their more attractive positions, as by inclining the fore part 

 of their bodies to the camera when it is desired to make the 

 shoulders look larger than they are. What I desire is that 

 other and inexpensive photographs should be procurable, which 

 shall be suitable for exact measurement. 



All that is asked for is a strip of hard level ground, on which 

 a rectangle is laid out of some 8 or 9 feet (say 100 inches) long 

 and 20 inches wide, and otherwise marked, and that the 

 camera shall be directed squarely towards a certain point in it, 

 as shown by the diagram (Fig. i). The horse is to be led to the 

 rectangle, and kept in it, taking care that all his feet stand within 

 its margins, that the cross-line at c lies clear of his front and 



Fig. 2. 



hind legs, and that the tip of every hoof and each corner of 

 the rectangle is visible from the camera. The appearance of 

 the horse standing upon the rectangle, as seen in the photo- 

 graph, would be this (Fig. 2), but the lines should be much 

 more delicate. The camera should be placed on a line strictly at 

 right angles to the side A E of the rectangle. When the horse's 

 head, as in Fig. 2, is on the side of A, the perpendicular in ques- 

 tion should be drawn from B, a point about a foot distant from 

 c, and on the same side as a. Then the image of the horse fits, 

 laterally, well into the field of the camera. When the horse's 

 head is intended to be on the side of B, a symmetrically situated 

 point D must be selected for the foot of the perpendicular on 

 which the camera is to stand. The distance of the latter from 

 the rectangle should be fully 20 feet. 



NO. 1 47 I. VOL. 57] 



The camera having an adjustable back, must be slightly 

 tilted downwards in order to get a good view of the feet, but 

 its back must be kept strictly vertical. It would be a decided 

 gain if the installation were so arranged, that the photograph 

 should contain means of judging whether the plate in the 

 camera had really been adjusted aright — that is, parallel to the 

 vertical planes passing through the long sides of the rectangle. 

 This can be accomplished by laying out the rectangle near the 

 base of a wall, with its long sides parallel to it ; then, driving 

 two nails into the wall in the same horizontal line, hang a long 

 string with weighted ends over them, the ends nearly touching 

 the ground. If the two nails are well placed, the line will 

 appear in the photograph, as running along the top and down 

 the two sides of it, close to the margin. Then, if the plate has 

 been rightly adjusted, the upper line in the picture will remain 

 parallel to the long sides of the rectangle, and the two weighted 

 ends will remain paiallel to one another. Otherwise there will 

 be convergence in one or both respects. The camera must be 

 high, in order to show well the position of the horse's feet ; 5 

 feet is perhaps the best height for it at a distance of 20 feet, then 

 the perspective foreshortening of the width of the rectangle is 

 such that its scale is one-quarter of that of its length. 



Relative measurements can be made with accuracy on photo- 

 graphs taken under these conditions, between any visible points 

 that are situated on the median plane, such as between the 

 withers and the lower part of the chest below them, between 

 the front and the back of the profile, and so on, but absolute 

 measurements cannot be made because the distance between 

 the camera and the median plane is not yet accurately deter- 

 mined. Much less can heights above the ground be measured, 

 for in order to do so it is necessary first to determine the 

 line at which the median plane of the horse intersects 

 the ground, because it is to this line that the vertical 

 measurements must be made. What is meant by " median 

 plane," is the imaginary plane which passes lengthways 

 and vertically through the spine of the animal, and which 

 serves generally as a plane of reference. It cuts the ground 

 half-way between the two pairs of hoofs, so that if the half-way 

 position between the tips of the fore hoofs be called s, and 

 that between those of the hind hoofs be T, then the intersection 

 of the median plane with the ground lies along the line s T. 

 For its accurate determination the animal should stand on hard 

 ground, and the hoofs be so disposed that at least the tips of 

 all four sjiall be visible from the camera, which must be well 

 elevated so as to look down on them, as already described. 



Beginning with the simplest case, namely that in which the 

 median plane of the horse is parallel to the long sides of the 

 rectangle, and- also parallel to the plate in the camera, we 

 are at once in a position to measure the height of the horse, its 

 depth of body wherever desired, and its length. For by pro- 

 longing s T in the photograph until it cuts the ends of the rect- 

 angle in s' and t', a length equal to that of the line s' t' drawn 

 on the median plane in any direction will correspond to the 

 length of 100 inches objectively. We begin by measuring the 

 lengths of s' t', and those of any other, say two, dimensions on 

 the photograph. Call their several measurements /, a, and h, 

 reckoning them according to the scale used throughout for that 

 purpose, whatever the value of the units of that scale may be. 

 Let X and y be the objective values of a and b, which have to 

 be found ; 



then s : 100 : : a : x and :: b:y 



100 100, 



X = —j-a y = , b, 



s s 



so the coefficient ^- being determined, serves to convert these 



s' 

 and all other measurements in this same median plane into 

 their objective values. Those persons who possess Crelle's 

 Multiplication Tables, can perform these little sums without 

 effort and with great rapidity. In Fig. 2 the real length of the 

 rectangle there represented happens to be only 80 inches. The 

 measurements on the diagram are as follow: — (i) s' = 21 '6 

 mm. ; (2) height of withers, 157 ; (3) height of lower side of 

 chest, vertically below, 8*9 ; (4) height of rump. i6'2 ; (5) ex- 

 treme length of body, 1 5 "5. Whence the coefficient = — - = 3 7, 



21 'o 



and the objective values are (2), 58"i inch, (2-3) 25-1 ; (4), 



Before considering the effect of obliquity the following object 

 tion must be disposed of. It may be said that the protuberan- 

 sides of the animal will prevent its true outline being visible 



