196 



NA TURE 



[August i i, 1923 



length was 6*48 in. The optical centre of the photo- 

 graph was known. In place of measuring the tilt, this 

 was determined from the photograph ; the joints in the 

 floor gave the principal vanishing point, and the vertical 

 lines the nadir point. If the distance between these 

 two points is Q, then the angle of tilt = \ sin~^2F/Q ; 

 in this case the tilt worked out at 37° 57'. The distance 

 of the parameter parallel and field centre from the 

 principal vanishing point is p. If an upright arith- 

 metical scale, measuring from the parameter parallel, 

 is drawn alongside the photograph, this will be repre- 

 sented on the plan by a harmonical scale. Or, by 

 computation, if h is the perpendicular distance, from 

 the horizon, of any point in the photograph, the distance 

 of the representation of this point, measured from and 

 at right angles to the parameter parallel, will be 

 p{p-h)lh. The intersection of parallels so obtained 

 with rays drawn through the " field centre " will give 

 the positions of the points on plan. 



So far, the system is essentially a point by point 

 method of plotting. It will, no doubt, often be found 

 of real value, and the neatness and simplicity of the 

 few calculations required recommend it, especially in 

 those cases in which the plate is considerably inclined 

 to the horizontal. It is not so convenient when the 

 plate is nearly horizontal, and it would be a mistake to 

 suppose that it enables photographs to be converted 

 into plans which could not have been so converted by 

 the methods used before its publication. The data 

 required remain the same. The method has the minor 

 disadvantage that the plotted plan will depend for its 

 scale on the height of the camera, so that a mosaic of 

 photographs will give a series of plans on different 

 scales. But this is not an important objection. 



Mr. Gordon's book is somewhat troublesome reading, 

 and the student will probably find the clearest explana- 

 tion of his methods and terminology in the chapter 

 entitled " Recapitulation." But the book contains 

 original matter and will take its place in the*; list of 

 authorities which those interested in the subject must 

 study. 



Mr. Gordon states in his preface that recent develop- 

 ments of methods of military reconnaissance have 

 given " an urgent call " for a generalised system ; he 

 writes of " the authentic rule which the soldiers of 

 1915 so urgently desiderated " ; he states that he 

 provides the solution of the mathematical problem 

 " that grievously vexed the soldiers during the years of 

 the Great War " ; and he indicates generally that, in 

 his opinion, the want of a knowledge of Taylor's rule, 

 rediscovered by him, prevented the survey staffs of 

 the armies from making effective use of air-photo- 

 graphs in the construction of military maps. This 

 does not do justice to the work of the survey staffs. 



NO. 2806, VOL. I 12] 



The writer of this notice is satisfied that if Mr. Gordon 

 book had l)een available during the War, it would ha- 

 made no material difference. Mr, Gordon provides 

 new method of plotting ; but several other thorough! > 

 sound methods were in use. 



No difficulty was, as a fact, experienced in convertin 

 air-photographs into plans by the methods actually 

 used. It is not the case that oblique photograph' 

 were avoided on account of any supposed difficult 

 in making use of them. This is a mistaken idea. 

 Photographs departing considerably from the hori- 

 zontal were, in general, avoided, because it was neci 

 sary to get vertically, or nearly vertically, over the 

 enemy, to find out what he was doing and to avoid 

 the interference of cover. The photographs so 

 obtained, although taken on plates that were only 

 inclined to the horizontal some 3° or 4° on the average, 

 were not treated as plans, but were converted into 

 plans by perfectly orthodox methods. There is an 

 admirable exposition of the matter by Lt. -Colonel 

 M. N. MacLeod entitled " Mapping from Air Photo- 

 graphs," published by H.M. Stationery Office. All 

 who are interested in the subject may be advised 

 to read this. They should also read Mr. Gordor 

 instructive book, and they will then see that there arc 

 several ways of killing this particular cat. 



Two other matters call for special comment, namely, '. 

 contouring from air-photographs and the use of a 

 vertical base. As to the first, the theoretical difficul- 

 ties are not formidable, but the practical difficulties 

 are, and neither Mr. Gordon nor any one else has yet 

 properly solved them. At present Mr. Gordon's sugges- 

 tion is as good as any, and that is to plot two maps 

 of the same piece of ground from two different positions 

 of the aeroplane, and determine the heights, point by 

 point, by means of the varying parallaxes. This agrees 

 with the advice of Lt.-Col. MacLeod, which is to prepare 

 prints of two photographs, separately taken, " rectified " 

 to^ chosen horizontal plane, and from one of them to 

 make a tracing which can be superimposed on the 

 other, for comparison of parallaxes. But even this 

 method will fail when the points are not visibly marked, 

 and would be inapplicable to the contouring of an 

 ordinary hill-side. Perhaps something might be done 

 by stereoscopic plotting from two parallel plates 

 simultaneously exposed from the extremities of the 

 wings of an aeroplane. 



With regard to the use of a vertical base, Mr. Gordon 

 points out that it is theoretically possible, given, in 

 the oblique photograph, a vertical line which has 

 three points marked on it, at known distances from 

 each other, to determine the nadir point on the photo- 

 graph, and, the optical centre being known, the tilt 

 and parameter parallel can be found. But the practical 



