w-k^i&dMA' Peijrpeettve f of BeQtpjeT By EA."Wrtgl)t;Arcl)irect: Cornell University Library The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924015333945 ARCHITECTURAL PERSPECTIVE FOR BEGINNERS. oo3srT-^ii!sriKrc3- Eleven Plates of Practical Examples, CONSIDERED WITH REFERENCE TO A STUDENT IN AN ARCHITECT'S OFFICE. F. A. WRIGHT, Architect. SECOND EDITIOX. NEW YORK: WILLIAM T. COMSTOCK, 23 WARREN STREET. 1890. Copyright BY F. A. WRIGHT 1885 PREFACE T N preparing the following plates on Perspective, I have mainly kept in mind that large class of *• draughtsmen who are obliged to be self-taught. Since the time when I myself was struggling to obtain from the best books to be had a practical knowledge of perspective, I have been more and more convinced that all the books were either inadequate or too complex, if considered solely with reference to the actual needs of an architect's office. It is all very well to discuss vanishing points, visual rays, parallel planes, and the phenomena relating to perspective, but what the student in an office wants is the solution of the- practical problem of making a perspective of a house. If a proper training through a good course of study at an architectural school was open to all, perspective would be easily acquired, the student having the advantage of teachers to consult, who would direct him how to apply the principles there taught to his practical needs. But there are comparatively few draughtsman who have the benefit of a school training; and knowledge of perspective is acquired in a hap-hazard way, more or less slow, according to the ability of the learner. I hope that this book may be of special assistance to draughtsmen who have but little time to give to perspective, and that the start they may obtain through its aid may lead them to study and learn more. All kinds of students, even those at schools, will find these practical examples of service by re- inforcing and making clear the application of theoretical principles. I have known of many students who, after a four-years course at an architectural school, found that they had considerable of this sort to learn after they had entered an office. I would like to state that the method of this book is the one that I have followed with my offiae pupils, and that the rapidity and ease with which they acquired facility to draw in perspective made me think that others could as easily be taught in the same way through a book. This method is to take the student at once to the practical problem; to show him how to put a given house in per- spective, and then to inculcate theories and principles by a progressive practice. The plates first appeared in Building, and their reception there, as I have learned through many letters, emboldens me to think that in book form, with much new matter, they will be welcome by many more. How to shade a perspective after it is drawn is also a hard question for the young draughtsman to work out. I have tried to help him a little in this respect, and if he models his work after some of the drawings herein given and referred to, he can not fail to acquire something that will eventually increase his value to his employer. U9 Broadway, N. Y. City. F. A. WEIGHT November Sji, ISSu PLATE I PLATE I exhibits the ground plan and two elevations of a small cottage, and the perspective of the same, with all its principal oon- p. . structive lines shown by dotted lines. The design is selected to J represent as simple a building as one is called upon to make a per- spective of in ordinary professional practice. Let the student not be alarmed by a seemuig confusion and multiplicity of lines shown in the plate. The whole matter is simple enough upon examination, and there will be no difficulty in separating the lines that are purely con- structional from those which are not. The plan and elevations are drawn on the plate for convenience, and explain themselves. An architectural perspective drawing is made by the aid of a per- spective diagram. The diagram is shown at the top of the plate. The _. first step to be taken is to make this diagram by the aid of Persoective ''^^ plans and elevations. Let us first, then, ascertain what _. it represents, and how it is'drawn. A perspective diagram is an outline drawing showing the plan of the building, with all the various openings, such as windows and doors, marked on it, and also showing the roof plan as well as all the various projections of the building beyond the planes of the main walls. It is, in fact, a plan of the building that shows all its external features which appear in the perspective. In this case, for convenience, it is drawn to the same scale as the plans and elevations. Draw it on a separate piece of Manilla paper of relative size shown. After the diagram is finished it is placed at an angle at the top of the board. The drawing is, of course, made in the usual manner, and not while the paper is in position. It wiU be noticed first that only the roof plan and projections of the two sides which appear in perspective are drawn. It should be carefully drawn in firm lead pencil lines (or ink, if too complex and apt to be marred by use). A careful study of the diagram shown in the plate, and the relation of the various points shown on it to the plan and elevations, is advised. This should be self-explana- tory to the student as to how it is made. The ground plan is represented by one line corresponding to the outside line._ The points a, b and c, on the diagram, are the points a', b' and c' on the ground plan. The lines joining these points represent the two principal vertical planes of the exterior which show in perspective, beyond which, or within which, the different projections are marked, and to which they are referred. The windows and door openings are indicated as at m n, the measurement of openings being taken between jambs or the inside lines of casings. The front porch, with its posts and steps, are laid off same as shown on ground plan, after which the roof plan is superimposed in proper relative position, with its projections beyond the main planes of the buUding all noted. The roof lines are taken from the elevations, and the lines representing it are the extreme projections. The plan of the chimney is shown in its proper posi- tion on main ridge. With our diagram now drawn so that the various points in it represent accurately by scale all the principal vertical lines and points of the two sides to be shown in perspective, we are now ready to take the next step. The piece of paper with diagram on it is placed at the top of the draw- ing board, at about the centre, in position as shown in plate. After it is _. tacked down, the horizontal right line Pi Pr, marked "picture p. . plane," is drawn, running clear across the board. For conveni- Plane ®°°® ^'^ taking heights, it is drawn touching the point b of the diagram. The line thus drawn represents a vertical plane upon which the picture is supposed to be drawn. We have next to select the station point, which is the position in which the observer is supposed to stand, and from which the object, represented q. . . by the diagram, is viewed. Drop perpendiculars to the picture p . . plane from outer extremes of the diagram, cutting it at Ii and i. About midway between these points draw the line C Ps perpen- dicular toP P. (Let P P stand for picture plane.) It would be best, in general, to draw this line cutting P P exactly midway between points h and Ij although it is not necessary, and not so drawn in the plate. On the line thus drawn we can establish our station point. It can be taken any- where on the line at a reasonable distance in front of P P. In the example it is taken about 61 feet in front of the nearest corner of the house. It Horizon. would be better to take it 100 or 150 feet away, according as the size of the board permits. Drive in a pin at Ps, which is our station point. The paper on which the perspective is to be drawn can now be tacked down. It can be placed anywhere between Ps and the picture plane. For convenience it is usually placed near Ps, at the bottom of the board, about as shown in the plate. Place it so that the line C Ps will fall across the middle of the paper. The horizon can now be drawn. This is a line drawn with T square, at right angles to C P». It can be taken anywhere between Ps and P P on the paper, the only limitation being that it must be located so that all the perspective will fall on the paper. In the plate it is taken about one-third the way up from the bottom of paper. Generally it is nearer the bottom — from one-fourth to one-eighth up, according to the height of the building. Draw it across the paper and extend it across the board on either side indefinitely. The next things to be determined are the vanishing points. From Ps draw lines parallel to a b and b c, cutting the P P at Pi and P^. The line P Pi is thus parallel to one set of horizontal lines in the diagram, and the line Ps Pr to the other set, at right angles to the first. Let fall perpendicu- lars from pi and P", cutting the horizon at Vi and V. All horizontal lines parallel to a b wiU vanish at Vi, and all parallel to b c at V. Drive in pins at these points. Drop a perpendicular from bj making it a firm, light pencil line. This is the nearest corner of the house, and it is in the P P. This being so, it can be used as a line of heights, and the distance of any point „^ .° . ° of the building above or below the horizon can be measured UB ainmg ^^ j^. ^^ ^j^^ savie scale that diagram is drawn to. In the plate Heig s. ^jjg elevations are so placed that the heights can be directly projected across to this line by means of the T square. This can be done in making any perspective if there is room enough between the vanishing points and the paper for the sheet containing the elevations to be tacked down. Draw the horizon across the elevations. In such a case the eleva- tions must, of course, be drawn to same scale as the diagram. If, as is apt to be the case, the elevations are not drawn at the same scale it is desirable to use for the diagram, heights must be measured from them and laid off by same scale as diagram on the line b 25. It onlyremains now to fix the height of tlie horizon itself. The horizon is on a level with the eye, and it is this consideration which determines its relative position on the elevations. If the observer is supposed to stand on a level represented by the grade line of front elevation, his eyes would strike the building 5 or 6 feet above that line when he looked directly at it. In the plate the horizon was taken a little over 5 feet above the grade line. After its position is located, it should be marked upon the elevations, unless they are placed alongside of perspective as in the plate, when it is only necessary to so place them that the horizon will fall across them at the required proper height. Careful study of the constructive lines should make clear how, with the data so far obtained, the perspective is drawn. The various points of the diagram are ail drawn down to the picture plane by visuaV ^ rays from Ps,*| From the points where the visual rays pierce erspec ive. p p^ perpendiculars are let fall as required for use in deter- mining points and lines in the perspective. AU the different points and vertical lines sought will be found on or coinciding with the perpendiculars thus drawn. The first lines drawn would be those from a and c, cutting P P at k and j. From k and j drop perpendiculars as shown, defining two corners of the house, the other corner coinciding with the line of heights already drawn. Next, lay off on b 25 the height of the belt course above and the water table beneath the horizon. Measure on the elevation from the horizon, or these heights can be projected directly across with T square from elevations. The points 26 and 2 7 represent the heights of the belt course and water table as laid off on the line of heights. Lines from these points are drawn to V' and VI as indicated, those portions of the lines falling within the vertical lines already drawn being the perspectives of the belt course and water table. In the same manner all the horizontal lines that lie in the main vertical planes of the house, such as the window sills and iintels can be directly obtained. The side gable shown on left-hand side of perspective is obtained as follows: The height is projected across cutting b 25 at o. From o an in- definite line is drawn to VI. The peak of a gable, whose plane lies in and coincides with the main plane of house wall, is somewhere on this line, and, as we know it also on the line r, obtained from diagram, it fol- lows that it must be at the intersection of these two lines, which point is found to be at 10. This point is represented on front elevation at 28, and on diagram at 12. The peak of the actual gable is represented on the ele- vation by the point 29, and on diagram by point 13. It can be seen that the point 10 which we have obtained is a point in the line which forms the ridge of side gable, and as the ridge line vanishes at V' we have only to draw an indefinite line through 10, and the intersection of this line with the line s will give us 14, which is the real gable peak. A visual ray from point 30 in diagram, which is where the side ridge pierces the plane of central roof, by its perpendicular let fall to intercept the perspective of the ridge, gives the point 17 of the valley line 17 18. _ The perspectives of the lines forming the main eaves have to be de- termined next. Lay oflf on line of heights the height of point 31 (so marked on both elevations). This height falls at point 7, and a line to V establishes in perspective the line of intersection of the roof and wall planes. A line to V| gives the point 32. Connecting the points 10 with the points 32 and 7, we have the perspective of a gable lying in the plane of the walls. But, the roof projects beyond the walls, and-it is necessary to find the point 9 (marked on the diagram and elevations 34). When we have found the correct height of this point it is only necessary to draw the eaves through it to Vr, and a line to VI establishes the point 11, which is on the line t at the height of the eaves. In the diagram b 8 represents the actual projection of the eaves. Measure the height of eaves on the line of heights. This height is indicated by point 2. Through 2 draw a line from VI. A visual ray from point 8 with its perpendicular intersects this at 3j and a line through 3 to Vf represents the correct height of eaves in perspective. A visual ray with perpendicular from point 34 in diagram establishes point 9, and a line to VI through 9 gives point 11. The main gable in front could be obtained in the same way that the side gable was, but, for convenience, and in order to show two methods of ar- riving at the same result, its height is found in another way. The student will find as he goes on that there are several methods of doing the same thing; and that the choice of the particular method to be employed is chiefly a matter of convenience. One method also serves as a check upon the other, and hence it is desirable to know them all. In the diagram the line marked " main ridge " is extended forward until it cuts the P P, and from this point a perpendicular is let fall, marked "auxiliary line of heights." All points which lie in the vertical plane of which this line is a vertical trace, can be laid ofif on it by scale (measured from the horizon), and their position directly determined by the intersections of lines drawn to VI with the lines from the corresponding points in the diagram. The height of the main gable is projected across at z, from which point a line is drawn to VI, stopping the line when it meets the line of side ridge. This is the line of main ridge, and the intersection of it and a line from gable peak in diagram gives the gable peak in perspective at once. The lines x and w give the vertical lines of front gable, and the height of the eaves is found in same manner as for side gable. The hip roof of porch is also determined by the aid of the auxiliary line of heights. The height is shown at point 22, through which a VI line is drawn until it intersects the vertical line v. The heights of gable win- dows are found in the same way as indicated. The chimney height, projected at y, is found by the intersection of a line to VI from y and the line q, the line q being the intersection of one of the chimney planes with the vertical plane through main ridge. The in- tersection of the other main plane of chimney with the roof is found in the same way from point 6. The main outline of the building being now obtained, the finishing lines, such as the gable verge boards, the outside lines of window casings, the window sash, the porch railing, brackets, etc., are put in by the aid of the eye alone. As an exercise, the student is recommended to draw the examples here given on a larger scale — say four feet to one inch. Take the point of sight 150 feet away from P P. This will add interest to the exer- cise, and make it the means to learn much more than would be learned if an absolute copy were made. Plafe 1 Qrduno plan . Jca/c o/ ^eef PLATE JI The diagram. PLATE I shows how the perspective of a small house can be obtained. In doing this something was learned concerning the perspective diagram, the picture plane, the station point, the horizon and the vanish- ing points. The method by which the perspective was drawn is applicable to all cases. A careful study of Plate I should give the student a fair insight into the practical principles of perspective — such an insight, in fact, as would enable him, even if he were to have no further teaching, to go on, and by applying the principles, ultimate^y to make himself familiar with the subject. Many students have started with less information and . have been self-taught by practice. Plate II exhibits a more elaborate example than the first — a large coun- try house being the subject. From lack of space but one elevation is Plate Si^sD. and the ground plan, by the aid of which the diagram was J J made, is omitted. The perspective itself is shown in outline, only such lines being drawn as are obtained directly from the elevations — the lines of construction. All the other lines, which, with those drawn, go to make up a finished perspective, are obtained by the aid of the eye. At least it is not necessary to obtain them from diagrams and elevations, although they might be accurately found in this way. In regard to this ' point more hereafter. As has been said, a perspective diagram could be made by the aid of the elevations if there were no plans. Although we have here but one eleva- tion, it will answer to show the student the relation between it and the diagram. Remember what it is desirable to show on the diagram — the roof plan of the building, and an outline plan with all the openings marked on it. It can be drawn to any scale, i. A> i> or larger, according to the size of picture wanted. For ordinary houses the -^j will be found a good scale, giving a perspective not too small, and large enough to exhibit plainly all the necessary detail and finish. The diagram of Plate II is drawn to the same scale as the eleva- tion — on a separate piece of paper and tacked down at the top of drawing board. The horizontal line marked " picture plane " (P P) is next to be drawn. First, a word as to its location. In Plate I it is drawn touching the near comer The Picture °^ *^® house. This was for convenience in taking measure- Plane ments, and in most cases this is its best location. But if, for any reason, it is desired to make the picture smaller, the P P can be located down towards the station point (Ps) as shown in Plate II. The nearer the P P is to Ps the smaller becomes the perspective and vice versa. If the P P was located above the diagram, as it might be, the size of the perspective would be very much increased. This is hardly ad- visable, as it is much better to draw the diagram to a large enough scale to permitVthe P P to be located below it. If drawn above, any error in projecting the various points of the diagram (by means of the visual rays) would be increased, while if it is drawn below, any such error is reduced. In Plate I, for convenience, the diagram was so placed as to have both sides of the house make equal angles with the P P, t. e., angles of 45°. This is not necessary nor advisable. A better result is generally obtained by showing more of one side than of the other, i. e., by placing the dia- gram so that one side makes an acute angle and the other an obtuse, as shown in Plate II. The diagram should be placed near the left-hand side of drawing board. The station point, from which the visual rays connecting the various points of the diagram with the P P are drawn, is established as described Station ^°^ Plate I. From the extreme right and left hand points of the Point diagi-am let fall perpendiculars cutting P P at a and 1). Midway between these points draw a line at right angles to P P, running to bottom of board, and on this line (100 feet from P P) a pin is put in to denote the station point (P»). After the paper upon which the perspective is to be drawn is tacked down (anywhere between Ps and P P, as convenience in working may seem best), the horizon is drawn extending across the drawing board. Note what was said concerning it in Plate I. Owing to the fact that the diagram is inclined to P P more to one side than the other, the vanishing points will be at unequal distances from the Horizon. „ . . . line on which Ps is located. Vf will be further away than p . . VI, hence the reason why the diagram was placed to the left. Vf and VI are found, as in Plate I, by drawing lines from Ps parallel to diagram lines representing the front and side vertical planes of the house. These lines cut P P at P' and Pl (outside the plate). Perpendiculars from P' and Pi to the horizon give Vr and Vi, both of which points lie outside the plate. All the horizontal lines lying in or parallel to the vertical plane of the longest side (front) of the house will vanish to Vr, and all those at right angles to these to V. In drawing any perspective, the heights of the various points in it above the horizon can all be obtained by the following method : Produce Measuring- ^^ ^^^ traces of vertical planes containing the lines vanish- T ■ ing to Vr and VI until they meet the P P. From the points where these various lines produced meet the P P, drop perpendiculars, and on these perpendiculars the heights can be laid oflE — on each perpendicular for the points in the plane it represents. Another way of stating the method is to consider the diagram made up of horizontal traces of vertical planes. By producing these traces until they meet P P, itself a horizontal trace of a vertical plane, and then draw- ing perpendiculars down from the intersections, we get the respective vertical traces of the planes represented in diagram. These vertical traces are measuring lines for all points lying in the planes they are the trace of. A clear understanding of this is important. The explanation hereinafter given under "the perspective," showing the way the method is put into practice, will perhaps give a clearer idea of the matter. Something was said concerning the convenience of locating the P P on the near comer of the diagram. The reason it is convenient is because by so doing you get the two principal vertical traces in one line, and that line is one of the lines of the perspective. This is what was done in Plate I, and when it is done all the heights can be taken on the vertical line so ob- tained. Some of the heights, however, were laid off on what was called an " auxiliary line of heights." What was there done for a few heights is done in Plate II for all the heights. First establish the position of the horizon on the elevations. The usual height is that of the observer's eye, 5 feet 6 inches above the ground. Draw a line across elevations at this height. All measure- mentSf are taken from the elevations (same,Bcale as diagram) measured above and below this horizon drawn on them. They are then laid off on the various lines of heights (measuring lines) from the horizon of the perspective. The first lines drawn in the perspective would naturally be the vertical lines representing the principal corners and angles of the house. Visual rays from Ps piercing P P at c, d, e, f, g, h, i, k, 1 and m are a few such lines. Perpendiculars dropped from' these points give the position of the lines in perspective. All other vertical lines and all the various points ' represented in the diagram are obtained in the same way, i. e., by drawing visual rays from Ps to the points, and then dropping perpendiculars from the points where the rays cut P P. Produce the two main planes represented in the diagram until they meet P P. There dropping perpendiculars gives measuring lines No. 1 and No. 3. On No. 3 we can lay off t u and t t below horizon, equal to dis- tances denoted by same letters on elevation. Lines to VI from n and t give the water table and ground line in the perspective as shown. On same measuring line lay off height of belt course t r, balcony floor t q, balcony railing t p, and main cornice t o, these heights all being equal to the vertical distances between points with same letters on elevation. The plane which was produced to give the measuring line No. 3 is a plane whose horizontal lines would vanish to Vi. Hence in laying off heights on No. 3 V is to Tdb, used to transfer them to the perspective.' AU the heights might be laid, off on No. 3 and then transferred to perspective by first using Vi and then Vr to get points in proper position in planes van- ishing to Vr. It is more convenient, however, to produce the other planes until they meet P P, and then erect measm-ing lines, as we have done. On No. 1 we can lay off (same way as above described), height of main The Perspective, gable at x, height of balcony gable at j, height of kitchen gable at z, and all other heights lying in main plane of house— windows, etc., and points not shown by construction lines on the perspective, but all of which the student will understand are obtained in the same way as described for the few that are so shown. On No. 3 the points 7 and 8 are the heights of the cornice above and the ground line below the horizon for the horizontal lines of the wing to V^ Points 15 and 16 on No. 9 are the same measurements for the lines to V. The horizontal lines on the corner that is cut off on this wing will vanish neither to V or VI. They are found by joining the intersections obtained from points 7, 8, 15 and 16. On No. 4 the height of the upper right hand gable is laid off at 6. The same height is also laid off on No. 7 at n. It is a good idea to get heights from different measuring lines for the same point. When the two agree it is nearly certain to be correct. On No. 5, 17 is the height of side gable. The height of the windows and base line of gable should also be taken on No. 5. Point 9 is the height of the chimney. Use VI to 10 and then V'. No. 3, besides being a measuring line for the main side plane of house, is also the line of measures for the veranda. As will be noted, the planes of veranda and side of house produced meet P P at same point. Hence the measuring lines coincide. Points 1,18 and v are heights taken on No. 3 and transferred by Vf to veranda. Measuring lines Nos. 6 and 8 serve for height of main gable (s), heights of kitchen chimney (11 and 12), kitchen cornice (13) and porch cornice (14). Do not forget that all these heights are laid off on the measuring lines at the same distances above horizon that the points are above the horizon marked on the elevation. The same points on the elevation are lettered with the same letters and figures. The center of any parallelogram is at the intersection of its diagonals. This is true for a perspective parallelogram, which, as we know, differs from a regular parallelogram in that two of its sides are not really parallel, but only appear so. For instance, in the parallelogram 2, 3, 4, 5, it is evident that the line 2 3 is not really parallel to 4 5. The vertical lines 2 4 and 3 5 are actually parallel. The parallelogram 2, 3, 4, 5 is the perspec- tive of a parallelogram, and the perspective of its center is found in the same way that we would find it if it were not in perspective. This is an Important thing to know, and capable of wide use in drawing perspec- tives. AU the gable peako could be found by diagonals without reference to the diagram . For instance, after the points 4 and 5 are determined, if the parallelogram is completed by erecting perpendiculars to V line from 17, and diagonals be drawn (3 4 and 2 5), a vertical line through their intersection will give the gable peak, coinciding of course with the line obtained from diagram. i Another application of the use of diagonals is shown for the veranda gable. Here the parallelogram is already formed, and we have only to draw the diagonals 19, 22 and 21, 20. In the same way we might find the other peaks, and many other points. Fig. 1 shows an application of the use of diagonals in finding the apex of a square tower or spire. Here the parallelogram used is formed on a vertical plane passing through opposite corners of the square. This plane passes through the apex ; hence, by drawing diagonals as shown, and erecting perpendicular, we obtain a line which contains the point we wish to find. Touching the further use of diagonals in drawing perspective more will be said later on. The perspective in Plate II. is shown in outline only. Only the linea which are obtained by the aid of the diagram are drawn. As regards the other lines necessary to fiU in and complete the drawing, they can be put in accurately enough by the aid of the eye from the data already fixed. Much depends on good judgment and ability to interpret the elevations, i. e., to know from them how the detail would appear— just what would be seen, and what hid by intervening projections. In the gables, the in- clined lines of the mouldings on the near sides {i. e. . sides nearest P P) would appear as two lines only, while on the far sides, four or more lines would show, as in Fig. 3. The finish of a window in masonry is shown at Fig. 3. If in wood, it would be the same, except that the outside line of the casing would be shown. The amount of detail or finish thus Bhowu in a perspective is in proportion to its size. The smaller it is the fewer finishing lines and detail. The point C in the perspective is the point where a line from the eye of the observer strikes it perpendicular to P P. It is called the centre of the picture. If you look at this point with one eye closed, perpendicular to the paper, and with the eye at a distance away from paper, equal to the distance of P from P P, you will see the perspective under precisely the conditions which we have assumed in the data. Looking at it from this one fixed point, and in one fixed direction, you may be surprised to find that everything appears more distinct and clearer than when you look at it in the ordinary way with both eyes. For an exercise the student had best take the elevations of some simple building at hand and try to put it in perspective by following the same steps thus far described. V PLATE III FOR our third example a chapel is taken. It has a round belfry tower and circular and pointed arch windows. These features give oppor- tunity to explain something concerning the practical manner of placing circles in perspective. It is desirable to show the perspective on as large a scale as possible to avoid confusion in the constructive lines. Hence it is made as large as the size of the plate will permit, and the elevations are omitted for lack of room. From the two previous plates it is hoped that the student is familiar enough with the method of obtaining heights and the manner in which they are transferred from the elevations to the perspective. It is also hoped that the matter of making the perspective diagi'am is fairly understood. One or two points, such as the relation of Ps to P P, and the location of the horizon, can be further dwelt on to advantage. It is partly a matter of judgment as to just how far in front of the picture plane it is best to- locate the station point. There are special Relation of ^^^^^ when it is best to take it not far away. The subject Ps to P P °' plate III may be instanced as one of them. The chapel stands in a high knoll of ground, and is ordinarily looked at from the lower level of the road near by. To conform to these condi- tions, and also because the building is thought to look more picturesque, Ps is taken only 64 feet from PP. A third and principal reason is be- cause a short distance gives greater emphasis and variation in showing the perspectives of the circle. In general, however, it may be stated that the greater the distance from Ps to P P the better the perspective. This dis- tance is limited by the size of the drawing board, and, although it is pos- sible to have it too great, still, with perspectives made from diagrams to the quarter, three-sixteenths, or even eighth scale, the best results will be attained when it is taken as far away as the size of the board will admit. As a general rule, the less the distance of Ps from P P the more distorted is the perspective. That is, the more inclined are the lines, and certain detail is more exaggerated. It can easily be seen that as Ps is moved toward P P, the distance to the vanishing points Vr and VI propor- tionately decreases, and hence the greater the inclination that lines to Vi" and VI from the same point make with each other. To avoid this acute pitch of the lines and consequent distortionate eflEect, Vf and VI must be far removed, and, as these distances depend wholly upon the location of Ps, hence the reason for taking -it in general as far from P P as possible. The plate shows the effect which the location of the horizon may have. The eye of the observer, looking horizontally, strikes the ground beneath Location of the building. This throws the building high up. The further Horizon down the horizon is "taken the higher up will the building be, and we see less of the roofs and inclined lines and sur- faces, especially those which incline toward the observer. This effect is what is known as fore-shortening. As the vanishing points of right lines are always on the horizon, it follows that the further down it is taken the more inclined are the lines drawn to the vanishing points. It is, there- fore, not a desirable thing, as a general rule, that the horizon should be taken very low. As the horizon is moved up, more and more of the roofs will appear. When it is midway between the vertical height of the per- spective there will be the minimum amount of fore-shortening of various inclined lines. When it is taken clear above the building the result is a bird's-eye view. It must not be forgotten that the horizon is located v/ith- out reference to the picture plane or vanishing points. The distances of the vanisliing points depend entirely upon the distance of Ps from P P, and have nothing to do with the horizon, i. e., they are the same distance away whether the horizon is low or high. As to the best position for the horizon, it will not do to lay down any rule. If it is desirable to represent the building on rising ground— and it very frequently gives by far the most picturesque result — the horizon should be placed low. Not so low perhaps as in the plate. More frequently the water table line or ground line is taken. If the building is viewed from the comparatively level ground on which it stands, the horizon will naturally fall from 4 to 6 feet above jhe ground line of the building. The diagram runs off the paper a little on the right. The visual rays from ps connecting all points in the diagram with P P, from whence PI t III they are perpendicularly let fall to their proper vertical place in ' the perspective, are not all drawn down as in the two previous plates. If this process is not thoroughly understood the student should refer back. Fig. 4 and Fig. 5 are not parts of the diagram. They are projections of the rose windows vfith their circumscribing octagons, shown where they are for convenience. The P P is made to pass through the corner of the two main vertical planes of the building produced. Mention has been made of the con- venience of so locating it. It will be found in most cases the best thing to do. If for any reason — such as desiring a smaller perspective — it is taken down in front of the diagram, it vrill be necessary to produce the main lines of the diagram, as fully described in Plate II. The perspective of a circle is the line in which P P intersects a cone of visual rays. The vertex of the cone is in the eye of the observer at Ps, and p . . its base is the circle itself, as represented in the diagram, of Circles According to the theory of conic sections, this line of inter- section will be a circle, ellipse, parabola or hyperbola. If P P cuts the cone parallel to the base, i. e., to the plane containing the circle, it will be a circle. If P P cuts the cone parallel to its axis, it will be an hyperbola ; if at an angle parallel to one of the rays forming the cone, it will be a parabola; and if at any other angle it will be an ellipse.* It is not proposed here to show how these may be mathematically de- termined. An example of the case where the perspective is a circle will be given in the next plate. Cases where the perspective is a parabola or hyperbola, are unusual, and can be disregarded. When they do occur they can be found by the same method used for the ellipse. In ordinary perspective the circle appears as an. ellipse, and the problem of putting a circle in perspective is to find the ellipse which represents it. An ellipse can of course be drawn with precision after its axes are known, and these can be determined in the perspective, but it is unnecessary to do it in this way. For all practical purposes the method hereinafter described will be found the best and most efficient way. This method simply consists in supposing a square or octagon circum- scribed about the circle. The centres of the sides of the square or octagon give at once four or eight points of the ellipse sought. K a line is drawn tangent to the sides and' i)assing through the poin,ts we wiU get the re- quired perspective with sufficient accuracy for all ordinary purposes. If greater accuracy should be desired — as might be the case if the perspect- ive was of unusual size, the number of sides of the circumscribing polygon can easily be increased, thus giving so many more points. Let us first take the rose window of the left hand gable. Fig. 4 is a projection of this window with its circumscribing square and octagon. The diagonals of the square bisect four sides of the octagon at Method of P°i''*^ where the circle is tangent, i.e., four points in the Putting Circles '^^'^°^®' '^""'^ °'^®^ ^"^^^ points are obtained from the in Perspective ^^^^^'^^^ ^^^ horizontal diameters of the circle. The P • octagon is projected into the diagram as shown. Fig. 4 is drawn where it is, for convenience in showing the relation of the various points to the diagram projection of it. It is not necessary to draw it in this position. It would naturally be drawn on the elevation. Bear in mind that the circumscribing square and octagon really have a vertical and horizontal position, the same as if drawn on the elevation. Visual rays cutting P P-at a, b, c and d, and thence let fall indefinitely into the gable (supposed to be already drawn), give the vertical lines of the circumscribing square and the vertical lines on which four of the octagon's corners must be. We have only now to determine the corre- sponding horizontal lines and complete what will be the perspective repre- sentation of the circumscribing octagon and square. On the line of meas- give ures the height are laid off at e, f, g, and h. lines from e and li to VI ♦There is one case in which PP may cut the cone and give a circle instead of an elhpse. This occurs when the base of the cone makes the same angle with the axis as P P does, but in a contrary direction. It is what is known as a mb-contrary section. the perspective horizontal lines of the top and bottom of the square. Lines from f and g determine the other corners of the octagon. It will be under- stood that the vertical heights laid off at e, f, g and h are equal to the hori- zontal projections in the diagram. That is, e f=g h=i j=k 1 and f g=j k. The circumscribing square is now in perspective, and the octagon can be drawn in perspective by joining the points where the vertical lines from c and d,and the horizontal lines from f and g, cut the sides of the square. The diagonals of the square being drawn, a vertical line and VI Une through their intersection, give the eight points referred to in Fig. 4 in pei-spective. A line drawn through these points and tangent to the sides of the octagon gives the perspective of the circle sought. The jamb or reveal of the window is represented in diagram by the lines m i and n 1, m n being the line of the window frame back of main vertical plane. The inside circle representing it is put in perspective by the same method as above described. It is hardly necessary in this case to use a circumscribing octagon. It can be drawn accurately enough from the circumscribing square. The dotted line is the part not showing, only the further side being seen. The dotted lines show clearly enough without further explanation how this second circle or jamb line is put in. Where the circle is a small one, or the drawing on a small scale, it is not necessary to use a circumscribing octagon. The square alone will answer in most cases, as in the small porch window. The octagon will answer for as large a window or other circle as occurs in ordinary practice. If the circle is of unusual size, or great accuracy was desired, a circumscribing polygon of 13 or 16 sides can be used, after the same manner as here de- scribed for the eight-sided polygon — the octagon. The rose window of the right hand gable has all the finish lines drawn in. After the second or jamb circle is drawn, its circumscribing square may be divided by its diagonals, and vertical and V' diameters. These lines help to determine the tracery. The center circle with its quatrefoil would be first put in, after which the other tracery can be drawn in free hand, the diameters of- the circle and diagonals of the square being the centre lines of the quatrefoils. Fig. 7 shows on a larger scale a method of determining the arch ring and voussoirs. There are two squares, one whose sides is equal to the diameter of the intrados and the other to the extrados. In other words, the arch ring is determined by circumscribing squares and octagons, and the ellipses are drawn as before described. By means of the diagonals and diameters we can divide the inside circle or intrados into 8 parts. It is then easy to sub-divide each of these eighths into the requisite number of arch stones or voussoirs. The voussoirs radiate from the perspective center of the circle — the inter- section of the diagonals — and, after being marked off on the intrados, the voussoir lines are so drawn by the aid of a pin. The voussoirs of one quad- rant being determined, the other quadrants could be determined from it by using the verticals and horizontal Vr lines, as shown in the Figure. A segment arch is put in perspective as shown by Fig. 8. The height of the spring line and the rise of the arch are laid off on the line of measures. _ . The center line of the opening is found by the aid of the ever 5 . useful diagonals. We now have three points in the arch through which its perspective representation may be drawn with sufficient accuracy. The voussoirs — brick in this case — will radiate to the perspective of the centre, as shown. In perspective, all radiating lines, of which these voussoir lines are ex- amples, will converge to the perspective of the point or center to which they converge in the elevations. It is oidy necessary to find the perspective of this point or center — always an easy matter to determine. To find the perspective of a pointed arch, a tangent polygon may be used, as in the porch door. If it is a large arch, it might be well to determine p . , . it by points in this way. Usually, however, three points are suf- °'" ^ ficient, as shown in the gable window next the porch. The '^'^ ' height of the spring line of the arch and its vertex can be laid off on the measuring line. Diagonals, as shown, give the centre line. With this data the curves of the two sides can be drawn with sufiScient accuracy. There is one thing to be borne in mind in drawing the curves of a pointed arch, and that is, that the near side, being a foreshortened line, appears as a sharper curve than the far side. The round belfry tower presents quite as difiicult a problem in the per- spective of circles as is likely to occur. The method employed in deter- „ . . , mining it is the same as that of the rose window. There the Horizon a ^.^^^^ ^^ .^ ^ vertical plane. Here the circles lie in hori- Circ es. cental planes. The first step is to obtain the vertical bound- aries of the cylinder forming the tower. The belfry level of the cylinder is represented in the diagram by the circle p, and the smaller diameter of the tower below by the circle q. Visual rays from Ps, drawn tangent to these circles, give, by perpendiculars let fall from their intersection with P P, the vertical boundaries r, 8, t and n. The projection of the belfry cornice is represented in the diagram by the circle o. Circumscribing squares and octagons are drawn on the diagram around the circles o, p and q, and the squares and octagons, lying in the same horizontal planes of the circles which they circumscribe, are then put in perspective. The height of the top line of cornice (represented by circle o) is laid off at y and the bottom line (circle p) at z. The square and octagon for each circle are then drawn as shown and the eight points of tangency found by the aid of the diag- onals of the square, t and 9, and the diameters x and w. Note that diag- onal Fig. 9 coincides with the line of measures, and that diagonal t is a horizontal line parallel to P P. Hence t in perspective is a T-square line. The points 1, '2, 3, 4, 6, 6, 7 and 8 are the points through which the circle o is drawn. Diameter w bisects the sides of square and octagon at points 2 and 7, and diameter x at points 3 and 6. The circles p and q of the belt course below are found in the same manner. All the circles are shown completed by dotted lines. In practice. it is, of course, only necessary to draw the portion that shows. In the cornice but two lines — top and bottom — are shown. Fig. 9 is a section of cornice, and fig. 6 shows how the intermediate lines would appear. These lines can be well enough put in freehand, the guiding lines of top and bottom being drawn. Fig. 10 is a section of belt course show- ing how belfry story projects. It will be well for the student to make himself familiar with the appear- ance of circles in perspective. To this end it is recommended that a copy of the plate on a large scale — ^three-eighths or one-half inch scale — be made. Change the point of view. Take Ps further away, and the horizon either higher or lower. Incline the diagram to PP so that it makes an acute and an obtuse angle— 60° and 40° say. Locate Ps on a center line to the right of the tower. These changes wiU make a different per- spective, and not being an absolute copy, the exercise will be a greater benefit and the means of learning more. After the student has become familiar with the general form and pecu- liarities of circles in perspective— from practice in accurately determining them — it will not be necessary to determine so many points. This is true of all things put in perspective. The more famUiar one becomes with the principles, the fewer points it is necessary to laboriously obtain from the diagram. The eye, trained by practice, after a while can safely take care of all minor points and details. jy*" £:i%i°". -■HorijAflL. -(ii .::m'.o" lo'y'^ PLATE IV. T HE perspective of the interior of any room is made by the aid of the same principles and rules that have been applied in the three previ- ous plates. It is presumed that the pupil has now an insight into these methods, such as to enable him to begin an intelligent study of the princi- ples, involved. In this book no more can be done than to indicate the direction which this study should take. The methods so far used may be briefly summarized as follows : The object to be put in perspective is represented by a diagram drawn to scale. The object represented thus, is seen through a vertical plane called the pic- ture plane. The intersection of the visual rays from the observer's eye to the various points of the diagram, with this plane, gives the perspective. If, while standing opposite a window, you could outline on the glass the, various objects seen without, the result would be a perspective delineation. The glass represents P P. The station point is where you stand. The ob- jects seen through the glass are represented by the diagram. This is a simple illustration of the whole subject. The methods we have been using are founded on similar data. From these data we derive rules which enable us to find the intersection of the visual rays cut by P P, and the result is a perspective. We have seen that the vanishing points of horizontal lines are found by drawing lines through the station point parallel to the lines as represented in Vanishinc- diagram. Perpendiculars to the horizon, dropped from the points Points ■w^here these parallel lines cut PP, give Vf and Vl. These are van- ishing points for horizontal lines. The vanishing points of any line or system of parallel lines are found in the same way, by drawing through Ps a line parallel to the line or lines whose vanishing point is to be found. Horizontal lines have their vanishing points in the horizon. All other lines vanish above or below the horizon, according to their inclination. It is seldom that it is necessary to obtain other than the vanishing points V' and VI, but it is well to remember that vanishing points for other than horizontal lines can be found. The raking lines of gables, for instance — those lines that are parallel to each other — vanish to a point, as well as horizontals. For the near side of the gable such lines would vanish above the horizon, and for the far side below. It can be seen that lines through Ps parallel to the diagram representation of such gable lines would cut P P the same distance from the center line of the perspective as V or V"", but some distance above or below the horizon. In order to determine how far above or below, construct a triangle having for its base the dis- tance from Ps to Pr or Pi, these last being the points where the horizontal lines to establish Vf and VI cut P P. On this line lay oil an angle equal to the angle of the gable. On the other end of the line erect a perpen- dicular, thus completing the triangle. The perpendicular side of triangle will give the height of the vanishing point above or below VI or V^, as the case may be. The perspective is drawn on P P revolved over until it coincides witn the plane of the paper. Remembering this, the general rule in regard to vanishing points would be as follows : The vanishing point of any system of lines is found by passing through Ps an element of that system, i. e., a parallel line. The point where it outs P P is the vanishing point of the system. All lines have vanishing points. This being so, it might be asked why vertical lines are not drawn to a vanishing point. Vertical lines are parallel to P P. Hence an element of the system through Ps would cut P P only at an infinite distance. The perspectives of vertical lines are, therefore, drawn parallel to themselves, their vanishing point being at an infinite distance. If P P was taken so that the vertical lines were inclined to it, they would have an actual vanishing point, the same as other inclined lines. All lines that are parallel to P P have their perspec- tives drawn parallel to themselves. When P P is taken parallel to one of the systems of horizontal lines, of which there are two in most architectural objects, the result is what is known as parallel or one-point perspective. The other set or system of lines is, of course, at right angles to P P. This is the case with the ordinary interior perspective. The observer looks at one end of the room, seeing that end and the two Parallel Perspective. adjacent sides of the room at the same time. Parallel perspective, then, shows three sides of the object, while in two-point perspective only two sides are shown. Plate IV shows two interior perspectives. The larger onfe is an example PI t IV °^ parallel perspective. Let us confine our attention to that at first, leaving the smaller one for after consideration. An interior perspective, exhibiting more detail, is generally drawn to a larger scale than an exterior. The diagram is therefore made large, as shown. In this case it represents the plan of a hall. P P is taken as shown, brought far enough forward so that all we want to show in perspective will lie behind it. Like in exterior work, it can be taken anyvirhere in front of Ps. It is most convenient to take it where shown and less liable to give rise to errors in drawing. It is sometimes taken coinciding with the side of the room opposite the observer, as shown in Fig. 11. In this case it is necessary to fix upon a line (a) parallel to P P, which will limit the portion of the room to be shown and form the outside boundary line of the picture. ps is taken outside of the room. For the reasons spoken of in the last article, the position of the observer should, in general, be as far away from the object as the size of the board will admit. In most cases of interior perspective Ps should be taken far outside of the room. In the example the room is supposed to be cut offl by P P, the fourth side being removed to enable the observer to look in. ps is located on a linfe, not running through the center of hall, as in Fig. 11, but a little to one side. Putting Ps on the center line gives a symmet- rical perspective. The vanishing lines from the side make equal angles with the horizon, and just as much of one side of the room is seen as the other. It is considered to give better results when more of one side is shown than the other, as pointed out in the last article. The lines f and gr subtend the greatest width of the picture. The sub- tending angle at Ps should not exceed 60°. The greatest height of the Field of P^*'''"^® *'s° should not subtend a greater angle than 60° It is Vision found that a greater angle than this leads to seetoing distortion of the near portion of the picture. This is another reason why Ps should be as far away as possible, for the further away it is the more will this limiting angle take in of the picture. This is shown in Fig. 18, where it is evident that as Ps is moved away from a given fixed position of P P the subtending angle of 60° at Ps will include a larger portion on P P. The dotted lines on the diagram show the visual rays from Ps drawn down to P P. Their intersections with P P estabKsh the position of the vertical lines in the picture the same as described before. The first lines drawn would be b and c dropped from the extreme right and left where the plaster line of the walls cut P P. The height of the ceiling is then laid off by same scale as diagram on either 1) or c, and the right lines e and d are drawn, e being the ceiling and d the floor line. The horizon is then laid off by scale 5 ft. 6 in. above d. The intersection of the horizon and the vertical through Ps gives the point V, which is the vanishing point for all lines at right angles to P P. Bearing in mind what -was said in a previous paragraph concerning vanishing points, it will be seen that an element of the system of horizontal lines at right angles to P P will pierce P P at V. All other horizontal lines being parallel to P P will be drawn with a T square. The heights are laid oflE by scale on b or c. P P becomes a plane of measures. The method of obtaining heights is the same as explained in the previous plates. The cornice and wainscot are laid off on c. The lines are drawn to V (by a pin at V) until they intersect the corner line h. The cornice and floor line of the end of the hall are then drawn with parallel T square lines, and the door and windows put in substantially as in eleva- tion. The intersection of the cornice lines with the pilaster gives the cornice lines over the arch. On a perpendicular let fall from i the riser of the stairs can be laid oflE from d, as also the height of the newels. In the picture, part of the stairs are shown that are in front of P P. It is all that part that falls within the field of vision. It was not shown in diagram for lack of room. The circles, it is to be noted, are put in with the compasses, this being the case when the perspective of a circle is a circle, spoken of in the last article. P P cuts the cone of rays, of which the circles is a base, parallel to the base, hence its intersection with the cone — the perspective of the circle — will be a circle. The upper half of the hall entrance door is shown opening out at an angle of 45°. Its horizontal lines are therefore inclined to P P, and their vanishing point is found in the usual way by passing a parallel line through Ps and dropping a perpendicular to the horizon from the point where P P is pierced. Its lines are drawn vanishing at V' The width of the floor boards is spaced off on d. Lines to V give the floor lines in perspective. The hall ceiling is divided oflE in squares as shown. These are laid oflE first by scale on e and vanished to V. The parallel lines of the squares might have been obtained by the aid of the diagram, but, in order to show another application of the methods of de- termining points by diagonals, they were obtained as follows : From k to 1 there are four spaces. Divide the distance n o (any convenient distance) into four equal spaces, and from the intersection of lines drawn to V with the diagonal m o, perpendiculars to the cornice establish the points of intersection of the limes s q and r. The line m o is the diagonal of a rectangle m n o p whose base p o represents in perspective the line k 1 of the diagram. The line q could also be determined by dropping a perpen- dicular from the intersection of the diagonals of m n o p. Do not forget that this rectangle is in perspective, two of its sides (m p and n o) being perpendicular, but the other sides (m n and o p) drawn to V. The method here applied is capable of the widest use for determining points in any perspective. The proposition governing its application may be stated as follows : If one side of a triangle be divided in any way, the adjacent sides may be divided into proportional parts by means of lines drawn parallel to these two sides and meeting on the third side. Another application of it is shown for determining the panels of the wainscot. Beginning at t, divisions are marked off on the bottom line d equal in number to the panels between P P and the door. Notice that t is in the plane of the wainscot. From t a V line is drawn to w, that being the point where the vertical line of door intersects it. Connecting w with u gives the third side of the triangle. This point u is assumed. The dis- tance t u is then laid off in equal divisions, as many as there are panels. The points of division from t to u are then drawn to the third side with the aid of vanishing point V, and from thence they are transferred to the adjacent side w t by means of parallel T-square lines. Perpendiculars from the points on w t thus obtained locate the panels in the perspective as shown. In case the divisions are not equal they can be laid off by The horizontal lines of the ceiling squares might have been found by means of the diagonal x y. The intersections of this diagonal and the V lines of the squares already drawn give the position of the lines r H and q. On the whole, this would be an easier method of obtaining them than the one before described. It is not necessary to go into any further explanation of the example of parallel perspective so far considered. It is a simple case, and easily solved by the aid of the methods given in the previous plates. The student who has comprehended the previous work ought to have no difficulty in seeing, with very little study, the relation of things in the perspective to each other and to the diagram. To learn the most from the example, it will be well to. make a similar perspective, or to attempt an entirely different one— following, of course, the same methods as here employed. The smaUer perspective is an example of two-point interior perspective, that is, a perspective with two vanishing points, Vr and VI. The diagram is shown above, from which it is seen that the sides of the room are in- clined to P P. (The P P of the other perspective was extended and used for this example also). Ps is way outside the room. The lines on the right-hand side vanish to VI, and left-hand side to V^. Where P P cuts the sides of the room, measuring lines are erected, upon which all the heights can be laid off by the same scale as diagram. It will be seen tliat the perspective is obtained in precisely the same way as an exterior. Where it is only desired to show a portion of a room, this is the best way to draw it, but when the full interior effect is desired, the room had best be drawn in parallel perspective. PLATE V. ' I 'HE phenomena relating to perspective are simple and hardly need to ■^ be pointed out at this stage. The four previous plates have been the , . means, perhaps, of making the student familiar with the most Perspective important phases of the subject. Still it will be well to note enomena. ^ ^^^ matters and restate a few important conclusions. 1. It is obvious that everything seems smaller when at a distance from the eye than when near by. The further a given object is from the eye the smaller is the visual angle it subtends, and hence in a rectangle the farther side occupies less space to the eye than the near side. The rectangle appears like a trapezoid. The sides seem to incline to«ach other, although it is koown that they are really at right angles. The appearance of things in perspective is different from the geometrical drawings we make of them, which, as we know, represent things as they really are. A pei-- spective drawing represents things as they appear. •i. Standing in the street or on a railway track, the main phenomena — that of vanishing lines and vanishing points, are easily verified. When looking in one fixed direction, the horizontal lines, as they recede, appear to converge to one point. Looking at a building cornerwise, so that its two principal sides are seen, its horizontal lines appear to converge to two- pomts, one on the right and one on the left. All parallel lines, whether horizontal or inclined, appear to converge to vanishing points. It only needs careful observation to discover these points of convergence and thus to verify the statements made all along in regard to vanishing points. 3. Many have perhaps observed the apparent convergence of horizontal lines to some distant points. Few have noticed where these points appear to be with reference to the lines. Standing in the street or room, or look- ing cornerwise at any object, one has only to look in a direction parallel to the lines in question, and the vanishing point will be found directly be- fore the eye. This is a matter that can be verified by observation. 4. When one looks at any object, such as a large building for instance, even though standing in one fixed position, the observer naturally moves his eyes about in their sockets, directing them by turns first on one por- tion of the edifice and then on another. Without moving the head one is able to look up and down, to the right and to the left, within a consider- able limit. In perspective the case is different. Different data have to be assumed. Here the observer is supposed to look in one direction only — that is, without turning the eyeballs. Furthermore, in perspective, but one eye is supposed'to be used. The reason for this is apparent when one looks through the window pane, observing both it and what is seen with- out. Looking first with both eyesand~then"with^one Bye'cloSedfitvwill be noted that the relation of the objects to the pane is different. With two eyes a distinct image is formed on the retina of each. But one image is conveyed to the brain, it is true, but in perspective, we can deal with but one image; the visual rays must center at one point, not two, and hence but one eye is to be used. 5. When we speak of one fixed direction in which the eye looks, this direction is understood to be the horizontal. A perspective can be made with the eye in any fixed direction. While this is theoretically true, practically an architect is rarely called upon to make a perspective related to other than the natural horizon. Hence the fixed direction is the normal one. A man stands erect normally in a vertical position. The normal direction of the eye is therefore at right angles to this position, that is, horizontal. 6. The natural horizon to which a perspective is referred is among the phenomena capable of being verified by observation. Looking at any ob- ject in which there are horizontal lines, one of these lines will appear to be a straight and level line, the others appearing to converge towards it. Observation cannot fail to detect this line which we call the horizon. A good illustration of it is the natural horizon, the apparent boundary line between the earth and sky. 7. The eye is always on a level with the horizon. We say that the hori- zon is a straight line passing through the eye. In perspective we can sup- pose the eye at any height. If the eye is high above the level of the ground, the horizon will be found just as high. This can be seen when one stands on top of a building or high hill and looks over the surrounding country. Standing in such a position and looking normally (i. e., hori- zontal), the dividing line of earth and sky wiU appear to be on a level with the eye, just the same as when standmg on the ground. Sitting or lying down on the ground, the horizon still appears on a level with the eye. 8. The vanishing points of all horizontal lines are always found on the horizon. This follows from (3) and (7). Fig, 13 will help to make this plain, and also explain why it is that any vanishing point is found by look- ing in a direction parallel to the lines whose vanishing point it is desired to find. In Fig. 13, Ps denotes the position of the eye. Ps H is the direction in which the eye looks. Lines parallel to Ps H are drawn above and below it. The perpendicular distance between the lines may represent any given object. ■ Suppose that a b, a' b', ii" b-* and a" b' represent different positions of a perpendicular object. As a b is moved away from P» to a'' b', etc., it is evident' that the visual angle subtending it (a P*- b) grows smaller and smaller. When a b is movfed to an infinite distance away from Ps, the visual angle becomes zero. The visual rays Ps a and Ps b will then coin- cide with Ps H, and the height a b will be reduced to a point on the line ps H, and hence the parallel lines, of which a b represents the distance between, will meet at that point. This is in accord with the mathe- matical proposition that parallel lines, when produced to an infinite dis- tance, will meet. Parallel lines appear to meet, however, at a measureable finite distance. P P may represent the plane of the picture seen edgewise. It can be seen from the figure that, as a b is moved away from Ps, the visual rays will enclose, where they cut P P, a smaller and smaller por- tion of it, until finally, when a b is an infinite distance away, the vi&nal rays coincide with Ps H and P P in a point. This point is the perspective of the vanishing point. The horizon being on a level with the eye. is seen endwise where Ps H cuts P P. It is evident that all lines drawn parallel to Ps H will meet it when pro- duced to an infinite distance, and that they will all meet it at the same -fioint. Take, for instance, the upper and lower lines represented by the distance c b. As c b is moved away the visual angle subtending it de- creases to zero the same as a b, and finally coinciding with Ps H and hence cutting P P in the same point as before. The visual angle wUl coincide with Ps H when the perpendicular distance is taken an infinite distance away, no matter what this distance may be. Hence all parallel horizontal lines appear to converge to one point on the horizon. The vanishing point will be always found on Ps H. As Fig. 13 mky stand for a general case — for. all cases, the reason a va^iishing point is always before the eye when the eye looks in a direction parallel to the lines, is easy to be seen. The statements made in regard to Fig. 13 may be verified by means of a long paper cylmder, see Fig. 14. Put in pins or straight pieces of wire, cutting the axis of the cylinder and parallel to each other. Looking through such a cylinder, taking the axis for the line of direction, the farther end will appear smaller than the near end. The pins further away from the eye seem smaller than the nearest ones. The longer the cylinder the smaller will the farther end of it appear with reference to the near end. It is easy to conceive of a cylinder of such a length that the farther end would appear practically as a point. If, before the cylinder is rolled, parallel lines be drawn on its inner surface, these lines will plainly appear to converge to a point directly before the eye, as in Fig. 15, which is an end view of the cylinder. The phenomena relating to the picture or perspective have been inci- dentally touched upon in the previous plates. A statement of them, and of the most important rules governing the simplest kind of perspective drawing, wiU, however, not be out of place. 1. The perspective is supposed to be drawn on a plane surface known as the picture plane (P P). It is supposed to be drawn so that if it was trans- parent, every point and line of the drawing would cover and coincide with the corresponding points and lines of the object represented, as seen from a given position, the station point (Ps). 3. The distance and direction of the perspective are taken upon a line passing through the station point, at right angles to the picture plane This line is known as the axis. We have spoken of the line of direction in which the eye looks, with which it is identical. In architectural per- spective the axis is usually horizontal, hence P P is vertical. Fig. 16. 3. The point P P nearest the eye (Ps) is the centre of the picture. See C, Fig. 16. C is always on the horizon. The distance Ps C is the length of the axis. It is generally hest that C should fall on a vertical line through the middle of the perspective. 4. Fig 16 is a perspective representation of P P with a perspective drawn on it, showing the relation of the axis to P P and to the horizon. The horizon of the drawing may be defined to be a straight line passing through C. 5. The perspective is supposed to be drawn by the aid of visual rays pro- ceeding from the object to the eye (Ps). Fig. 17 is a perspective represen- tation of P P, P« and the oject, showing their relation to each other. The intersection of P P with the visual rays from a' b' c' d' gives the per- spective a b c tl. 6. The perspective of a point is where the visual ray from that point cuts P P. The perspective of the point a', Fig. 17, for example, is where the ray Ps a' cuts P P, which is at a. The perspective of a line is where a plane of visual rays from the line intersects P P. The plane of rays P« b' a' intersects P P in the line a b. The perspective of any line may be obtained by finding the perspective of points in the line. For a straight Ime it is only necessary to find the perspective of two of its points. 7. In order to make a perspective drawing three things must be assumed or given: (a) the object in its dimensions and position; (b) the picture plane by the position of its horizontal trace and (c) the station point with the length of the axis. Fig. 18 shows the relation of jihese three things and the resultant perspective. The perspective is affected by a change of any one of these three things with respect to the other two. It is a matter of judg- ment for each particular case as to just how these data shall be assumed. Experience alone can teach the student as to how they shall- be assumed to give the best resultant perspective. The manner of locating the horizon and obtaining the heights has been so fully explained in the previous arti- cles that it is uonecessary to again speak of these points. 8. Lines in the object that are parallel to P P, whatever their direction, have their perspective drawn parallel to themselves. See Fig. 17. Hence, vertical lines are drawn vertical in perspective. 9. All lines not parallel to P P are drawn converging to a point. We know from observation that the real lines seem to meet at a point called the van- ishing point. The perspective representations of the lines do meet at a point, which point is the perspective of their vanishing point. 10. Lastly, with reference to the phenomena of the perspective it is only necessary to restate the rule in regard to vanishing points: The vanishing point of any line or system of lines is found by passing through Ps a line parallel to the line or lines in question. The vanishing point desired will be found where this line pierces P P. In the lower part of plate V an example of bird's-eye perspective is given. A perspective of this kind is useful in showing a group of buildings where Bird's-eve '* ^® desirable to show them all in proper relation to one Persoective another. Where an ordinary point of view could, only show part of the building or buildings, a high horizon is used so as to bring them all in the field of view from one point. The draw- ing is made in precisely the same way as any ordinary perspecti'-e. from which it differs only in having the horizon located very high. The norizon in this case was 38 feet above the ground. Where there are so many gables as in this example it is worth while to find the vanishing points for the gable lines. It is not absolutely necessary to obtain other than V' and VI, but it serves as an excellent check to the errors one is liable to make in working from a diagram. The gable lines inclining' 40 the right, -have their vanishing point at V", which, as will be noted, lies in vertical over V^ The height of V" above Vf can be found as explained in the last plate. The gable lines inclining to the left will vanish to V™, which will be found in a vertical line over V, the same height as Vn. oj J table on^-~~~-_ PLATE VI. THE shadows cast upon an ordinary exterior perspective are generally of a simple nature, and to the student who knows how to project them c, J on an elevation the problem of finding their perspectives will be easy. The shadows can be first projected on the elevations and then transferred to the perspective in the same manner as the perspectives of other lines. In an interior view, where there are cross lights and reflected lights, the problem of shades and shadows is complicated, and not easy of accurate solution. It is not intended here to more than touch of these matters. The subject of shades and shadows, although closely related to perspective drawing, will best "be made the object of special study, if accurate and scientific results are desired. To the beginner it is only necessary to point out the way in which constant study and observation will enable him to become familiar with all the aspects and conditions that arise in common practice. In exterior perspective the most picturesque disposition of light and shade is obtained when the sun is assumed as shining on one side of the object, leaving the other side in shade. . It ia-.coiwenient to assume the sun in P P. The rays of light will then be cast across the paper in parallel lines, and the shadows of vertical lines on horizontal planes wiU be hori- zontal T-square lines. Shadows, projected easily enough on the elevation in light, can be cast to conform to this condition, and, then transferred to the perspective. It is only necessary to assume the horizontal trace of the- sun's rays at an angle equal to the angle of P P and the side in light. In -discussing the examples which foUow, it will be taken for granted that the student has a preliminary knowledge of obtaining shadows on the elevations. The part of the object away from the light is said to be in shade. That part exposed to the light is said to be in light, or the bright side. Anything Shades ■''■hich intercepts the light and prevents its falling on the bright side causes a shadow to be thrown on that side, the intercepting surface becoming the bright side. In shading a perspective there is to be borne in mind one difference between a shade and a shadow. A shade is lighter in tone than a shadow. There is not the same depth to a shade as to a shadow. There are exceptions to this, as when the shadow is cast by an object whose bright side is darker than the surface on which the shadow is thrown. In such a case the shadow might not be as dark as the shade. In perspectives shaded in colors many of the shadows are no darker than the shaded side, the distinction between the shade and shadow being made plain by the difference iti color. In perspectives not shaded in- colors, however, it is weU to follow the general rule to lay in all shadows in deeper tones than the corresponding shades. Notice the shadows in plate VI. In these days, when so many drawings are reproduced by the various processes of photoilithography , it ia necessary for the draughtsman to know LithoBTanh ^°^ *° shade with the pen. The perspectives that appear .^jj. , from week to week in the different architectural journals can be studied to advantage by the student, and much knowledge can be thus obtained as to different methods of rendering. Drawings to be reproduced should be made on smooth white paper, such as bristol board, or Whatman's hot-pressed drawing paper. Vellum tracing paper and tracing linen are also good materials, especially the first-named. I have also had good clear lithographs made from drawings on Manilla paper. The drawing pen should be sharp and the ink thoroughly black. Writing pens should have stiff nibs — EsterbrooWs 303 pen is the best for fine work. The Spencerian commercial is an admirable pen for sketching and laying in shadows. The lines drawn should be clean, sharp and firm. Avoid making gray lines caused by thin ink or weak handling of the pen. Drawings can be reproduced full-size if they are very clearly rendered. The best results are, however, obtained when the original drawing is made from one-third to three times as large as it is desired to have the lithograph drawing. As soon as the beginner has learned to draw an outline perspective he is naturally anxious to take the next step and be able to give the drawing Shadine- with ®°™® ^°^ °^ * finish. It is not likely that he is well Riirht Lines acquainted with sketching by the aid of the common pen, and it is more than likely that his first efforts will be made with the right-line or drawing pen, an instrument upon which he has learned to rely. While the best results are not attainable by too great a use of the right-line pen, it ifi well for the beginner to commence with it. After he has learned to lay on shades expressing different degrees of light and darkness by it, he will do well to gain by constant practice a knowledge of the more difiicult free-hand rendering characteristic of the best archi- tectural sketching. Which of the two sides of a house shall be shown in light is the first question to be decided. Generally it is best to show the main front of the house in light and the side in shade. In some cases, however, the side elevation may be the most important, and it may be that a more striking disposition of shadows can be obtained by putting the front in shade. In other cases, of which an example will be given later on, a good result is obtained by putting both sides in light. He is the best draughtsman who has good judgment in determining what will be the most effective way to shade a given example. In the plate a typical manner of shading has been followed. That is, the front is placed in light and the side in shade. The first thing to be shaded will naturally be the roofs. The beginner is perhaps accustomed to lining in the roofs in elevations. As a general rule, they present the darkest shade of the building. The lines are first drawn on the light side. Draw with light, firm lines, as shown. The lines are paced (by the eye) to represent the shingle courses. The roof lines on the sides in shade are next drawn in with heavier lines than those on the light side, thereby producing a shaded effect in the simplest manner possible. In the same way the lines of the clapboards and wall shingles can be drawn. Give light lines for the sides in light and corresponding heavy lines on the sides in shade. For the sides in light it is not necessary that all of the courses should be drawn. It will be better to draw them broken, as by so doing a monotonous appearance is avoided. In right-line shading, when the shades are laid on so evenly, the result is apt to have a mechan- ical and monotonous appearance. The beginner can avoid this stumbling- block only as he has artistic discrimination and is able to conceive at the beginning a fair idea of what he wishes to make the picture. The perspective is now ready to receive the shade lines. We will leave the roofs alone, as they are sufficiently shaded, but the clapboard and shingles will need some stronger shading. The underlying principle that ought to prevail in laying on all shades is that the shade lines should be subordinate to the lines indicating the outline of the perspective and the construction of^the-buil4ijjg.... That is to say, the shade lines should not be so strong as the lines indicating the shingle and clapboard courses, for instance. In fact, they should be the lightest of all the lines used. Besides this, their direction should be such as not to interfere or coincide with the other lines as far as possible. For instance, the shade lines for the shingles and clapboards should not be horizontal, because if they were it might be difficult to tell whether they did not represent brick work. The shade lines, therefore, are drawn vertical — finer lines than the rest, as shown. Part of the side is left incomplete, to show the stage of the work before the shade lines were put on. Notice that in the gable, where there are mainly diagonal and vertical lines, the shade lines are laid in horizontal. At first glance you do not see the shade lines at all. The front door is shaded in the same way. The gable verge board is shaded by a light broken free-hand line, and the brick chimneys and foundations are ruled in with horizontal lines. The shade lines all being ruled in, the shadows next deiuand attention. The outlines should first be marked off in firm lead-pencil lines. All the lines of construction, such as the lines of the roof and clapboard lines that are embraced within the shadow, should now be gone over again with the pen and made heavier — heavier than the corresponding lines of the sides in shade even. Or these lines may be made heavy in the first place, if the shadows are projected before any shades are laid on. The shadow lines are now drawn in. Take notice that they are heavier than the correspond- ing shade lines. The shadow lines on roof are drawn parallel to roof slant or rake, and the shadows on clapboards and shingles are expressed in vertical lines, in conformity to the same principles referred to in regard to shade lines. In some places the shadows are not completed, in order to better show the different stages of the work. A shadow, like a shade, should be transparent. That is, no matter how much depth it has, or how dark and heavy are the lines that make it, we should still be able to see under it or through it the surfaces it falls on. After the shadow is lined in, the clapboards, for instance, can be seen through it, as, for instance, under the porch. The porch shadow is emphasized by crossed diagonal shadow lines. The right-line pen may now be laid aside and the drawing completed with the common pen. The windows are the principal thing. As a rule, they are shown with window curtains, in order to give variety. The window spaces should give the impression of being the darkest portions and deepest shades of the drawing; but it is not necessary, nevertheless, to put them in solid black. The drawing might be touched up here and there to advantage, in order to relieve its mechanical effect; but, as our object is merely to point out the rudiments of shading, it wiU be best to introduce nothing at this stage that might prove confusing. The beginner had best lose no time in acquiring a facility to use and firmness in handling the common pen. To this end it will be well to practice making free-hand straight lines of different degrees of heaviness, until the hand can be depended upon to a degree of rapidity and certainty as to the result. The best work is free-hand work. The most artistic and attractive form in which a perspective can be finished is that of a sketch done in free-hand, when everything is expressed in the fewest possible lines. Some practice of this sort is indicated on the plate. First draw horizontal lines, as at I, of all degrees of thickness, and wider and closer together, until your hand acquires a freedom and certainty. Pass on to 2 for further pi-actice. At 3 notice the three shades obtained from each set of lines, and also at 4, which may stand for the maximum effect of shading by cross-hatching. Clapboard and shingle courses are frequently put in, as at 5, with a d'ouble line. When practising these lessons draw much longer and larger lines than are shown. The lines numbered 6 show a desirable effect easily obtained by a wavy line. No. 7 is good practice, commencing with a fine line and gradually making it heavier. A lot of lines put in very rapidly is shown at 8 and 10, the latter without taking the pen off the paper. These are useful shadings. It is good practice to make lines, as at 9, first of one thickness and then of another. Free Hand or Drawing Pen, which ? Shades and Shadows in Brick and Stone. It may not seem so apparent to all, but the superiority of free-hand work to the mechanical drawing pen is attempted to be shown in the drawings of the two dormers. The upper drawing took longer to make, and the result is sharp, clear, and decidedly stiff. Let no one take this as a model. The lower dormer was outlined in lead pencil and then put in free-hand. Its lines are just as clear as the other, but somehow there is an added interest. It seems more natural, and, instead of being hard and precise, it has a soft effect. It will be well for the student to thoroughly appreciate the difference between these two drawings, for on this much of his success as a draughtsman will depend. The lower half of the plate is taken up with a drawing that the student would do well to copy on a larger scale. Such a drawing had best be on Bristol board. Draw very carefully in firm hard lead pencil first, and then do not touch a ruling pen, but finish it entirely in free-hand, as it is in the plate. Notice first, that the side of the octagon on which the light shines direct has the liglftest shade of all. Notice also that the vertical corner lines are not drawn, the corner being defined by the shade lines being of different thicknesses. This is a very effective method of shading. To be successful it has to be carefully done, each line being brought clear up to the edge. Notice next, that the brick work below the sill course is laid in with thicker lines than that above, and that the brick window jambs are thicker still. Bearing in mind all these differences, line in all the brick. The stone work is next laid in with character as indicated, and the drawing finished as shown, which 'tequires no specj^l explanation. The shaded sides of the arch joints and soffit and of the recess wall are laid in next, and the deep shadow under the arch is the last and most carefully drawn work of the whole. Notice that the shadow is made deeper at the top by simply increasing the thickness of the lines, thus giving it great depth, and that, notwithstanding the heavy shadow on the doorway and window, all the construction lines are clearly seen. There is nothing obscure in the shadow. The stone steps, being of darker stone than the rest, are shaded with vertical lines. The windows are put in simply with wavy vertical lines, giving the sheeny effect of glass. PLATE VII. T Caution. HE remaining five plates are intended to present to the beginner actual examples of perspectives that may answer as types to be followed. ,, . , Some of them are facsimile reproductions from architec- Model Persoectives tural periodicals, and are representative of the best draughtsmen of the time. They have been selected to show as many different styles of shading and finishing as was possible. The merits and defects of each will be pointed out, and the method to be fol- lowed to produce similar results will be sufficiently indicated to give the student a clear idea of the work shown. Although it is only by copying the work of the best draughtsmen that the unaided beginner can hope to advance, yet he should be careful not to entirely adopt the style or mannerisms of another. All this work should prove merely a training that will ultimately fit him to express himself in his own way. Individuality is not to be sought for, for itself, but if one really strives to see things from the inside, after qualifying by experience, h3 cannot fail in the end to acquire a certain style of his own. The drawing at the top of Plate VII is a very excellent exam- ple of a style much in vogue now, and one easy for the beginner to acquire, because it involves so little shading. Plate VI erspec ive presented examples of what might be called the orthodox with both , J ^ ,. sides in lie-ht metfiod. On the whole the most artistic results are at- tainable when there is both shade and shadow, but it is also the raost difBcult rendering to succeed in. Let the student not be dis- couraged by this, remembering that most good things do not come with- out effort. The drawing above referred to was evidently inked in with the ruling pen. The lines of the roof and those indicating the shingles and clapboards were put in free hand. The shadow under the porch was ruled in. Notice, if you please, the following points : First, the slightly wavy character of the free hand lines, especially in the roof. Too much of this would be bad, but a little of it redeems the drawing from what would otherwise be a monotonous effect. Next notice that some of the shingle and clapboard lines are fine double lines. This also tends to relieve the drawing. The third and most important thing to notice is the treat- ment of the w 'ndows. They are strongly put in and yet they are not wholly unrelieved blackness — a common fault in some otherwise good drawings. It must not be forgotten that in shading a window it is the glass that is to be represented. This has its sheen and bright points and reflections of light, and hence needs relief. One of the windows in the drawing is shown open, and the distinction between the glass and the open space is well marked. It may be well for the beginner to practice first in this kind of rendering and to return to Plate VI later. After the building is shaded one naturally wishes to finish the picture by means of foregrounds, foliage and sky. This book was at first only intended to carry the student through outline perspective, ""s-S^ 3.n jj^^ g^g jj. jj^g been thought best to glance at shading the Foreground. ^ , , . , perspective, something must also be said concerning the other things that go to make up the complete picture. These, however, can only be alluded to. There is room for a difference of opinion as to whether an architectural perspective drawing should be a picture from the artist's point of view, or, first of all, a picture of the building. If we adopt the first view, that every drawing shall be made an artistic picture at all hazards — that this shall be its chief requisite— we run against di£flculties. An architect is not primarily an artist, and, although he is a better architect for having artistic perceptions, yet, because his art is the outcome of scientific knowledge, and because the more his artistic nature gains the ascendancy the more he fails as a builder (architect means chief builder), it will hardly do to expect from him the same kind of drawing that a regular picture- maker would produce. An artist, in making a picture of a building, studies, not to display the building itself in its architectural character, but to subordinate it to the general effect. Besides this, even if an architect has inborn artistic faculties of appreciation, he is seldom able to take the time to get the power of expression in a technical way that an artist has. Hence it would seem wisest to bend our energies, first of all, to making our perspectives clear representations of the design and to subordinate everything to a proper presentation of the architecture involved. If we can make our drawings pictures in the artistic sense at the same time, why so much the better, but let us not forget that the perspective is made, not as an adjunct of a picture, but as the chief thing in the picture, and that a certain amount of architectural detail has to be shown. This can be made artistic, but it must be architectural. If the beginner is entirely without natural ability to sketch foliage, he cannot do better than copy the work of others until he acquires something of a vocabulary for himself. Observe the means that good draughtsmen use to indicate trees and grass and skies. Best of all, sketch from nature yourself all kinds of trees and bits of foregrounds that may be worked in as settings for your perspectives. Do not work on a tree with indefinite ideas, but try to make a difference between your oaks and maples and elms. Observe the characteristics of trees, and try to draw them by their characteristics. The impossible and conventional trees of some draughts- men often spoil a well-executed building. The drawing of " Eedcote " illustrates well what has been said about the difference between an architect's picture and an artist's picture. It is a very happy combination. The building is strongly set forth, and while the setting is subordinate, the whole makes a picture that is agreeable and artistic. The lettering can also be commended for its straightforward simplicity. An architectural sketch may be complete or incomplete. When some special architectural feature is desired to be shown, Mr. Kirby's charming « . -^ , . drawing of part of a country house (lower part of Plate Architectural ^^„, „ ■< \ f Sketches. ' °^^^^ ^ good model. This is done in free hand over a carefully laid-out perspective in lead pencil. The foliage is not quite equal to the rendering of the building, and might have been omitted altogether with advantage. The best part of the drawing is the way the light is managed in the casement windows. Notice the shadows of the sash bars on the curtains. Plate MI ■L iu]|i. f y f" ^'^^^^- >-^*asit^^;;^ PLATE VIII AFTER the student has had considerable practice in making perspectives by the aid of a diagram, it will be well to practice sketching in perspective, free hand. Take Ranid *^® elevations and work out the perspective by the eye. Of course this Sketching- requires practice and abUity to see from the elevation the exact appear- ance of things, but when acquired it will be found a rare and useful accomplishment. When drawing such a perspective consider your horizon and vanish- ing points, just the same as when drawing an exact perspective — that is, bear these constantly in mind. The gable and tower peaks, and the centres of arches, and other openings, can be found by the aid of diagonals, and by the aid of triangles the windows and other openings may be accurately laid out. Eapid sketching is good, both for the architect and for his client; for the architect, because he can design better by making such perspective studies, both of the building as a whole and in its detail, and for the client, because he gets what he might otherwise not get for lack of time — a fair idea of the actual building. The outline sketch at the top of Plate VII is a reproduction from drawing actually made in a great hurry, in order to give a rough idea of how the building would look. The original drawing is about 16 inches wide. It was made with a commer- cial pen, and represents about two hours' work. There is some little shading under the arches, but with this exception the building is in outline only; and yet sufScient is shown to indicate the character of the building. The clever drawing at the bottom of the plate by R. Norman Shaw, one of the most noted of English draughtsmen, is a good example of an artistic architectural picture. , At the same time the drawing is prominently architectural. Every , ... , . thing is clearly managed to heighten the interesting features of p., a quaint house. The student will do well to take this drawing as a model of what a perspective may be made. The pen work is fine, and very little of the building is not touched, but nevertheless, the building with the sun shining on both sides is swathed in light. This sunny effect on the building is heightened by the dark foreground and trees with the light middle distance. The building is outlined with a ruling pen, and then carefully finished in a most painstaking manner with a fine pen. The gradations in the shading are fine and well studied. Notice especially the fine rendering of brick work on the corner tower and side of house, and the careful manner in which the shadows are laid in — with fine lines, and yet very strong. This drawing was evidently not made in a hurry, and the draughtsman has made every moment tell to a good result. It is almost a pity that this drawing could not be presented three times as large as our book will allow, for many of its striking good points are lost in the small scale. Plate W ^cw iw^. -ewn^ ^..v.c, HOUSE RECENTLY ERECTED AT HARROW WEALD FOB r .CaaOAU.R.* R . NORMAN SHAW. R C. ARCH' PLATE IX. Interior Perspectives. 'X'HE drawing at the top of Plate IX, although very much reduced from the original, will show with suflBcient clearness the method of shading employed by Mr. R. Norman Shaw for an interior per- spective. An interior can be made in free hand outline with very little shading, but instead of showing an example of this kind of an interior, it has been thought best to take two good examples of highly finished interiors. It must not be thought that the examples here presented comprise all the most desirable styles of rendering. There are as many styles as there are good draughtsmen. Consult the current architectural periodicals, and many drawings will be found offering ideas of technique that the perceptive student will grasp at once. In a note at the end of this book will be found a list of a few good drawings that, if the student is able to obtain, will prove advantageous. The interior, showing the fireplace, like all of Mr. Shaw's drawings, is very carefully done. It was inked in with a ruling pen in light lines and finished with a fine pen. Almost the entire surface of the drawing has been worked over with fine, even lines. These lines are delicate and yet firm. There is no indecision about them. The draughtsman evidently knew just what he was going to do before he put his pen to paper. One thing that helps this kind of shading and, indeed, any shading, is the absolute precision of the drawing in outline. The perspective is so well drawn (notice the mouldings and all the small detail) that, without a line of shading and without ref- erence to the interest of the design, it would be a masterpiece. Notice the rendering of the stone work over the arch-vertical lines and horizontal lines, and yet an evenness of tone characteristic of cut stone. The joints of the masonry are worth study. They give us the effect of square cut ashlar work, and yet they are very uneven lines— wider at some places than others, and chipped at the angles. The wall paper is laid in with a background of fine horizontal lines, over which the pattern of the paper is drawn. Notice finally the shading of the mouldings and the fine, sharp shadows which give accent to all the mouldings, especially those of the arch. All these small points go to make the whole a very successful drawing, and it is only by considering them that a like result can be ob- tained. Mention has been made of the drawing in outline as a good perspective. Notice that it is a perspective with two vanishing points, and made in the same way as an exterior. Where it is only wished to show The _ two sides of a room or of a single portion of it, such as a Perspective , . . , , ^ j, itself mantel-piece, two-point perspective is always best. In fact, it is always best to use two vanishing points instead of one when it can be done. One-point or parallel perspective gives rise to dis- torted effects more easily than two-point. Mr. Colcutt's admirable interior may well stand as a model for draughts- men to follow. There is nothing but praise to be said of it. Considered as a perspective from a well chosen point of view, or as a wonderfully good bit of shading, it is equally praiseworthy. And yet there is nothing striking about it. Take note that it is not always the most striking effects of light and shade that are to be aimed at. A subdued effect, especially in an interior, is far more artistic. In regard to the perspective, note that it is a two-point perspective, and yet three sides of the room are shown. This is a very happy idea. Vr is right in the drawing to the right side, just above the stair landing. The sides of the room vanish to this point in same manner as if this was the only vanishing point, as in parallel perspective. VI is a long way off on the left. The lines in parallel perspective that would be drawn with a T square vanish to VI. This arrangement is reached by slightly inclining the end of the room to the picture plane. The vanishing points are then found and the whole perspective is made in the same way as any exterior perspective. In regard to the shading, the same remarks made concerning Mr. Shaw's drawings will fully apply. The only special thing to notice is the effect of a polished surface produced in the wood work by delicate shadings, and bright lights on certain portions — the mouldings for instance. A Model Drawing. Plate K INTERIOR OF HOUSE , PARIS . t.e.collcutt archt PLATE X. T^RAWINGS made purposely for publication do not fairly represent the every-day work of an architect. The sketches hurriedly made in competition, and to show- to a client, may not be quite so finished. In order to give the student an idea of or- dinary work, a reproduction is made from a drawing that was used in a competition. The two trees are rather too symmetrically placed perhaps, still the whole effect is good. With reference to the shading here employed, it is only necessary to state that the perspective was drawn in lead pencil and then entirely finished free hand. Notice that the vertical corners are mainly indicated by shading without vertical lines. Shading But the chief thing to which attention should be given is the manage- without Shadows ™6nt of shadows. The sun is supposed to be full in front of the building so that the sides of the building would be in shade, and yet no shadows cast by the sides. Shadows are cast directly behind the part that is in light. Notice the shadow of the chimney. This is a rapid way to get a good effect. The result is apt to be more striking than when light is shown on both sides. PLATE XL A BEAUTIFUL design and drawing is here presented, not as offering a distinct model for the slnident to follow, but as an evidence of the great diversity of means a skillful draughtsman can employ. To make such a drawing as this requires mature experience and a precision of workmanship and confidence in the final result that can only come after long practice. Notice particularly any portion of the render- ing, and it will be seen that there are lines hatched in diagonally, vertically and horizon- tally that bear no relation to the orthodox shading to which we have been confining our attention. Then note how very full of light the drawing is. There are strong shadows and shaded sides, and yet on these shaded sides there is as much light in places as on the light sides. These are not things that the student ought to rashly attempt. It is only a master that can ignore anything and everything in detail and produce such an artistic whole that one never stops to enquire how it was done. Notice how the foliage is man- aged to help the building, especially the large tree over the centre of the building. Notice also that, notwithstanding the rendering, tiles and brick and wood are clearly distinguishable. Finally note the unusually dark windows. There is much food f oi reflection in all this. s o o -i < J- X * P z = O OB ^ X u 0) z o o > 3 ui NOTES. 1. For those who have the adrantage of using the architectural papers, I would point out the f ollowiag drawings as offering hints and suggestions on shading : Building News, April 28> 1883. " Shingle-side House." The interiors here shown (evidently drawn from a photograph) are in the best style of Mr. Maurice B. Adams, one of the most noted of English draughtsmen. The Architect, July 5, 1879. Interior "St. Mary's Church, Speenham- land." G. E. Street, R.A., Architect. Suilding News, June 27, 1879. "The Council Chamber, Wakefield Town Hall." T. E. Collcutt, Architect. American Architect, December 11, 1880. " Cottage at Far Eockaway." Bassett Jones, Architect. A very fresh drawing, with good foliage. Building News, October 7, 1881. " House at Battle.'' W. H. Oakley, Architect. This is as strong a piece of work as I have been able to find. Notice the shading on the roof. American Architect, July 31, 1880. "Cain's College." Mr. Water- house, Architect. A fine example of sketching. Building News, January 5, 1883. " Bridlington Priory Church." J. O. Scott, Architect. Building News, January 7, 1881. Prize drawing by Fredrio Miller. The finest outline perspective I have ever seen. Building News, April 1, 1881. Interior "Church St. Matthias." " " Sept. 80, " "East Gate House." " " Nov. 11, " " The Shakspere Dining-room Set." " " Aug. 13, " "Entrance Hall, St. Peter's Hospital." Building News, Aug. 13, 1881, " Hall of the Commandery, Worcester." Building News, December 15, 1883. "House at Datchet." Ernest Newton, Architect. A drawing very good in composition, from an artist's point of view. The curve of the road and the shadow of the trees falling across it form a fine effect and one that can be tried by anybody. Building News, June 9, 1883. Interior. By J. Nicholl, Architect. « " Sept. 8, 1883. "The Lady Chapel, S. Mary Overie." The drawings in the American Architect for the last few years, signed by D. A. Gregg, are excellent examples of clear shading. Building, Octoher, 1883. "Fairview." Rossiter & Wright, Architects. '' August, 1884. House at Netherwood. Rossiter & Wright, Architects. These two drawings show the picturesque effect of a low horizon. Two drawings by Rossiter & Wright in the American Architect, one October 27, 1883, and the other March I, 1884, serve to further illus- trate the style of work practiced by the author. 2. Architectural Picture-Making •with Pen and Ink, by Benjamin Linfoot, Architect, will give the student some excellent ideas. It is specially good on foliage and skies, and presents models of backgrounds for settings which every draughtsman will appreciate. 3. I most heartilycommend as the fullest a.nd best treatise on Perspec- tive, Prof. Ware's Modem Perspective. I do not consider it sufiBLcient for the beginner, or else this book had not been presented. But, for the stu- dent who wishes to look into the subject more fully. Prof. Ware's book is excellent. I wish that all draughtsmen might find the time to study this book.