TRANSACTIONS OF THE SECTIONS. 9 



second of these was, that the line joining any two points along parallel lines, 

 assumed at an equal distance from the line to which both are perpendicular, formed 

 right angles with each of the parallel lines. The author then went through the 

 series of geometrical proofs, which would, however, be unsuited to our report, con- 

 cluding with the proof of the twelfth axiom of Euclid. 



Models to illustrate a new Method of teaching Perspective. By H. R. Twining. 

 The object of this communication is to explain the principles of perspective in 

 such a manner as may enable those who draw to distribute their objects not only in 

 a correct manner, but in one agreeable to the eye. The method aiFords an intermediary 

 step between those rules which are demonstrated by diagrams in the usual treatises, 

 and those appearances which characterize natural objects themselves. The chief dif- 

 ficulty in enabling an audience to follow out the principles of perspective when applied 

 to solid objects is, that every individual sees these from a different position ; so that 

 such an explanation of the effect observed as is adapted to one individual cannot suit 

 another. Mr. Twining's method aims at overcoming this difficulty by placing an 

 image (with which each individual is supposed to identify himself) in the exact spot 

 which the observer ought to occupy, and which serves to mark the true focus of the 

 picture. 



Light, Heat, Electricity, Magnetism, 



On various Phenomena of Refraction through Semi-Lenses producing Anomalies 

 in the Illusion of Stereoscopic Images. By A. Claudet, F.R.S. 



The paper had for its object to explain the cause of the illusion of curvature given 

 to pictures representing flat surfaces, when examined in the refracting or semilen- 

 ticular stereoscope. The author showed that all vertical lines seen through prisms 

 or semi-lenses are bent, presenting their concave side to the thin edge of the prism, 

 and as the two photographic pictures are bent in the same manner and by the same 

 cause, the inevitable result of their coalescence in the stereoscope is a concave sur- 

 face produced by the necessity of converging the optic axes more to unite the ends and 

 less to unite the centres of the two curved lines ; more convergence giving the illusion 

 of nearer distance, and less convergence of further distance. The only means to 

 avoid this defect is to examine the two pictures in order to employ the centre of 

 the lenses, which do not bend straight lines ; but as the centre does not refract 

 laterally the two images, their coincidence cannot take place without placing the 

 optical axis in such a position that they are nearly parallel, as if we were looking 

 at the moon, or a very distant object. This is an operation not very easy at the first 

 attempt, but which a little practice will teach us to perform. Persons capable 

 of using such a stereoscope will see the pictures more perfect, and all objects in 

 their natural shape. — Mr. Claudet presented to the Meeting a stereoscope made 

 on this principle, and many of the members present could see perfectly well with 

 it. The author explained the cause of another defect which is very often noticed 

 in examining stereoscopic pictures, viz. that the subject seems in some cases to 

 come out of the openings of the mountings, and in some others to recede from 

 behind, — this last effect being more favourable and more artistic. Mr. Claudet 

 recommended photographers when mounting their pictures to take care that the 

 opening should have their correspondent vertical sides less distant than any two 

 correspondent points of the first plane of the pictures, which could be easily done 

 by means of a pair of compasses, measuring those respective distances. To 

 illustrate the phsenomenon of vertical lines, bent by prisms, forming by coalescence 

 concave surfaces, Mr. Claudet stated that if holding in each hand one prism, the 

 two prisms having their thin edges towards each other, we look at the window from 

 the opposite end of the room, we see first two windows with their vertical Unes 



k 



