ALBERT R. MANN LI»RARY AT ■ . , CORNELL UNIVERSITY Cornell University Library SB 416.F28 Horticultural buildings.Their construct! 3 1924 002 825 598 Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924002825598 HORTICULTURAL BUILDINGS. HORTICULTURAL BUILDINGS. Their Construction, Heating, Interior Fittings, &c., with remarks on some of the principles involved and their application. Mith J 2$ Illusttiations. By f.-"^ a: f a w k e s. NEW EDITION. LONDON : SWAN SONNENSCHEIN, Le BAS & LOWREY, PATERNOSTER SQUARE. S. COWAN AND CO., STRATHMORB PRINTING WORKS. PREFACE. Up to the present date no book has existed, from which a gentleman could obtain, in a complete, concise form, reliable information to assist him in deciding what garden-structures would best suit his requirements; nor one in which an architect could see just those constructional and mechanical points which should be decided by the horticulturist, or in which a gardener could find details beyond his province, but with which he should be acquainted. The mechanics of horticulture has failed to receive that at- tention from the architectural and horticultural press, which the importance of the subject has warranted. While a vast amount of knowledge regarding plants, flowers, and fruit has been disseminated weekly, but little has been written upon the proper construction and arrangement of the buildings required for their cultivation. The system adopted in this volume has been determined with regard to clearness and ease in reference. It will be noticed that, after general remarks, some of the elementary principles of astronomy and the sun's rays are introduced; their bearing on glass roofs, aspect, site, &c., discussed ; general hints on growing houses of all descriptions, conservatories, combinations of houses and subsidiary buildings included ; de- tails of the various parts of houses, interior fittings, necessary adjuncts, &c., added ; and notes on the general principles of heating, the apparatus, and its parts, brought prominently for- ward. Finally, some information regarding meteorological in- struments, legal points and insurance is appended. A SI PREFACE. Illustrations and tables are given where necessary. For the sake of perspicuity, technicalities are avoided as much as possible. In all remarks, the word "house" or " houses " refers to horticultural buildings. Dwelling houses are not meant unless so specified. Any information, bearing even indirectly upon mechanical construction, is included ; but anything having reference to the cultivation of plants is omitted, as it is not thought desirable, in the present volume, to trespass upon the province of the gardener proper. It may sometimes be noticed that the same information is given more than once. This, however, is only the case when mere cross references are undesirable. Different articles and sections necessarily overlap one another, and such overlapping is frequently found advantageous. If the present volume renders the private gentleman, the architect or the gardener any assistance, and supplies a neces- sary link between the architectural and horticultural interests : if the architectural or horticultural press is stimulated to a more extended development of its functions : and if, by a combina- tion of these results, horticulture is benefited, the objects I have in view will be attained. Since the first edition was published a portion of the book has been rewritten, the information corrected up to date, a number of fresh illustrations engraved, and a great deal of letterpress added. My thanks are due to those firms and gentlemen mentioned in various parts of the book who have so kindly placed infor- mation at my disposal, and allowed me, for my own purposes, to illustrate their productions. F. A. FAWKES. Mansion House Buildings, Lojidon. INTRODUCTORY REMARKS. I CONSIDER that space is not wasted in discussing elementary points of astronomy, meteorology, the solar rays, &c., in con- nection with the erection of glass-houses. It will be found, not only that some of the physical sciences form an integral portion of the mechanics of horticulture, but that without a knowledge of some of the laws of Nature, it is impossible to properly build and fit up such an apparently simple structure as a glass-house. When once correct principles are known, an application of those principles becomes comparatively easy. I intend therefore to devote several pages to this matter. The following notes may be useful to amateurs and others, who, perhaps for the first time in their lives, contemplate the cultivation of plants under glass. First of all, have a clear idea what and how much you wish to grow. This point is most frequently lost sight of altogether. Whether it is a problem which requires some little trouble and thought to work out, or whether it is considered that other and more easily-defined limits are preferable, is immaterial, for cer- tainly, although it should form the first question, it is too often allowed to merge into the next question — " What is the limit of money which it is desirable to expend upon the proposed buildings?" Then is there a danger that a mistake of an important character will be made by the question resolving it- self into, " What is the maximum space I can cover for a fixed 4 INTRODUCTORY REMARKS. or a minimum sum?" If once this idea is carried out, a liberal crop of error is sown, which will inevitably result in reaping an extensive harvest of trouble. If you have no other way of limiting your proposed horticultural buildings than a monetary one, by all means use it, but do not abuse it. On the other hand, do not let your ideas be too extensive for your pocket, either as regards first cost or working expenses. If you formu- late your plans and they prove too costly, cut them down in dimensions rather than efficiency. What is worth doing at all is worth doing well, and a house thirty feet long, properly con- structed and efficiently worked, is far preferable to a corres- ponding house, double the length, costing the same sum, but scamped in construction and only receiving half the necessary subsequent attention. Having, say, only a hazy idea of what you want to grow, make yourself master of all the facts of the case. Investigate the local and incidental circumstances, such as aspect, situation, soil, drainage, best description of building or buildings for the purposes required, most suitable dimensions, heating apparatus, interior fittings, as well as all the attendant conditions which go to determine what you will build. By all means consult your gardener. Remember that he will have to use the pro- posed glass work, and will probably not only have a keen notion of what will best suit the desired purposes, but will take greater interest in structures, in the planning of which he has had a voice. At the same time, beware of crotchets. Gar- deners are only human. So far these remarks will apply chiefly to growing-houses. If the contemplated building or buildings are destined for show- plants rather than growing them, then the mind must be con- centrated upon points of architecture and art, decoration, and fittings, upon the beautiful in mechanics and materials, which will harmonise with the beautiful in nature. Reference should. HISTORICAL. 5 be made to the article on " Show-houses," in which this phase of the subject is treated in detail. HISTORICAL. The Chinese have for some time been acquainted with green- houses, but how far back their knowledge extends is not known. According to Mr. Loudon, the first green-house of which we have any record was erected about 1619 at Heidelberg, by Solomon de Caus, architect and engineer to the Elector Pala- tine. This green-house was originally constructed to shelter orange trees. Between this time and the end of the 17 th century, myrtles, sweet bays and heaths, as well as the orange tribe, were sheltered by houses having windows only on one of^ n. more perpendicular sides ; for such purposes a large amount of light was not then considered advisable. A green-house in the Apothecaries' garden at Chelsea was mentioned by Ray in 1684. It was not, however, till about the beginning of the i8th century that the desirability became apparent of rendering a large amount of the sun's rays avail- able, and glazed roofs became at all general. The following precis of instructions given in the " Encyclo- paedia Britannica" published in 1810, for the construction of a conservatory or green-house, may be considered amusing at the present date : — " No glass roof, but a seed-room over. Cornice about I Sin. deep. Lower brickwork about i8in. high. Piers of masonry 2ft. 6in. thick in front and i8in. thick at back. At every 7ft., intermediate masonry about 7ft. high filled in by glass work and shutters to fold behind them. Floor raised 2 to 3 feet above adjoining ground, trestles of wood inside with forms upon them as stages." 6 INTRODUCTORY REMARKS. When the duty was taken off glass, horticultural buildings, which before were a decided luxury, now became almost a necessity. So far as can be gathered, the first artificial heat employed in houses in England, was obtained by placing burning embers in holes in the floor. The antiquity of hot-water heating is beyond a doubt, for Castell, in his " Illustrations of the Villas of the Ancients," gives us drawings of the mode of heating the Therms of Rome, as described by Seneca. By these we see that the baths were heated by passing the water through a coil of brass pipes in the fire. Thus, before the Christian era, those principles were applied, by which, at the present day, we carry out hot-water heating. The commercial results of the discovery of America and of the passage to India by the Cape of Good Hope gave an im- petus to gardening, and it became necessary to erect green- houses for the shelter and cultivation of the rich and rare fruits and flowers received from tropical countries. It will thus be seen that the artificial treatment of plants, which has culminated in the latest improvements connected with horticultural buildings, may be divided into three distinct stages. The first was clearly the desire to protect, during the winter, pleasing and delicate evergreens from frost. Following this, in natural sequence, came the desire to perfect the means for the acclimatization of plants, &c., not indigenous to our soil and climate. Then, later still, came the wish to anticipate, by artificial means, the natural ripening and blooming seasons of fruits and flowers. . The first stage gave mere protection, in which glass played a very unimportant part, the walls being formed of brick or stone with windows at intervals, which formed the only means for admitting light. The second stage gave us buildings in which the walls had a SOCIAL. 7 little more window space in them : a certain proportion of the roof was glazed and some means were used for artificial heat- ing, either by putting a stove in the building (hence the term " stove " — building in which a stove is placed), or by placing burning embers in a receptacle underneath the building (hence the old term for a stove, " hypocaustum " — from Greek words meaning " under" and " to burn"). The third stage gave us forcing-houses proper — minimum obstruction to the rays of light by the maximum amount of glass in the roof and sides; means for the delicate manipulation of heat, light, and ventilation ; and various appliances, such as plunging beds with "bottom heat," as well as atmospheric heat, &c., &c. SOCIAL. The social advantages of a green-house are great. We know that "contact with, and a proper contemplation of, God's works, has a refining influence on mankind," so that a due ap- preciation of a garden will be morally elevating. A green- house enables both the contact and contemplation to be inten- sified, therefore a person cannot but be ennobled who thoroughly appreciates and properly uses a green-house. To watch in a glass-house the growth of a pretty little helpless plant, to pro- mote its development amidst adverse external circumstances, to shield it from cold, to protect it from the sun's scorching rays, to deliver it from its insect persecutors, to feed it, all these go far to touch in the human mind a mysterious chord of sympathy for the little plant, to soften the temper, and to es- tablish a fascination which is equally powerful to the aged couple who are approaching their golden wedding day, to the happy bridal pair in their new home, or to the grandchild of twelve summers. 8 INTRODUCTORY REMARKS. Fruits, as well as flowers, are doubly valuable, and they have to us a very peculiar flavour (which the most educated palate of a connoisseur would fail to detect) if they are produced under our own care, and in our own hot-house. Pleasure and fascination are increased if we take our microscope into our green-house, go deeper into the mysteries of plant life, and examine the exquisite beauty, order, and method in the hitherto invisible. The wonderful epidermis and storaates of pelar- goniums, scalariform ducts of ferns, spiral vessels of roses, pollen, the numerous insects which find their whole world on a plant, and the thousands of strange phases of organic life re- vealed - by the microscope, will afford food for delight and reflection during many spare hours. A pessimist will be trans- formed into an optimist when he sees that the world is no longer a " barren andhowling wilderness," but replete with unthought of beautiesj Our green-house will thus become a reservoir of /interest and information, in fact, a scientific kaleidoscope, [having an almost inexhaustible series of delightful combinations. ' Although the more extensive the horticultural buildings, the greater will be the tangible results which may be obtained from them, it not unfrequently happens, that the pleasure experienced by one person of enjoying, for the first time, the luxury of a little green-house only a few feet square, is as great as the in- terest taken by another person in an extensive range of hot- houses, requiring the attention of several men, and producing fruit and flowers, of every description, all the year round. If a contemplation of the beautiful, the good, and the true has a refining influence, then, apart from plant life, we should strive, to the utmost extent of our ability, to render our con- servatory a thing of beauty, and, therefore, a joy for ever. The judicious selection and disposition of plants and flowers, a chair or two, a curtain, and a little old china, may turn a conven- tional conservatory into an artistic floral reception room. ASTRONOMICAL. 9 There can be little doubt that philosophers are right when they say that every man should not only exist and live, but should have means for recreation and rest apart from sleep. A con- servatory properly built, and thoughtfully and artistically fur- nished, affords an excellent opportunity for healthful rest and recreation. Upon social, hygienic, moral, and even religious grounds, horticultural buildings may justly claim advantages of a solid and superior character. ASTRONOMICAL. In discussing green-house arrangements, our first difficulty is with the seasons. So long as we have winter and summer, cold and warmth, long and short days, it is much more satisfactory to talk about our green-houses and conservatories after we have a clear notion of some of the reasons for these various changes. We experience cold in winter and heat in summer, not because we are nearer the sun in summer than in winter (in point of fact we are nearest the sun about January ist, and farthest about July ist), but by reason of the " inclination of the earth's axis to the plane of the ecliptic." This will be clear from the following : — For the sake of argument we will suppose Fig. I.— The earth's axis at right angles to the plane of the ecliptic. that the earth spins round on its axis A B (Fig. i), and that lO ASTRONOMICAL. C D is the plane of the ecliptic (a globe-like orbit or path round the sun), also that E F, the equator, is coincident with this ecliptic. It is obvious that, to a spectator, on any part of the circumference E F, the sun would seem to describe a complete semi-circle in the heavens, rising due east, passing across the zenith and setting due west. To such a spectator this semi-circle would never vary. Similarly, the apparent path traversed by the sun, to a spectator on any other part of our globe, would never vary. Now, inasmuch as the relative position of the earth and sun is not as shown by Fig. i, but as shown by Fig. 2, in which the axis A B of the earth is Fig. 2. — The earth's axis inclined to the plane of the ecliptic. inclined about 23° 28' to the plane of its orbit C D (E F being the plane of the equator), it follows that the apparent path of the sun to a spectator varies throughout the year. The sun's declination is the angular distance between the equatorial tone and the ecliptic. As the planes C D and E F in Fig. 2, where they touch the earth, both represent circles on the earth's surface, it will be seen that these two circles intersect eath other at two points which are called the equinoxes (from aquus, equal, and nox, night). One is called the vernal equinox (about March 21st), and the other the autumnal equinox (about September 21st). When the sun reaches his furthest point, north of the equator, it is called the summer solstice. This occurs June 21st, and is of course the day on which the sun is the greatest length of time above our horizon. ASTRONOMICAL. II When the sun reaches his lowest point south of the equator, it is called the winter solstice. This occurs December 21st, -.^- Fig. 3. — Angle of sun's maximum Fig. 4. — Angle of sun's maximum altitude during the longest day. altitude during the shortest day. when the sun remains the shortest time above the horizon. The maximum height the sun attains at the summer solstice is about 62 ° (61° 57' 15"), and at the winter solstice about 15° (15° 2' 45") above the southern horizon. These angles are represented in Figs. 3 and 4. Here we have one of the reasons for our climatic changes. For, supposing a pencil of solar rays (these are practically parallel) to strike a flat surface when directed at the angle shown in Figs. 3 and 5, and another pencil of rays, of the same diameter, to strike the surface at the angle shown in Figs. 4 and 6, it is obvious that the latter will cover a much larger area than the former. Rays, therefore. Fig. 5. — Maximum inclination of pencil of solar rays during the longest day. Fig. 6. — Maximum inclination of pencil of solar rays during the shortest day. falling on A B, Fig. 5 (other things being equal) will feel much warmer, because more concentrated, than rays on C D, Fig. 6, 12 ASTRONOMICAL. It is clear that if the obliquity of rays were increased until the surface was in the same plane as the rays, the surface would cease to receive the rays at all. The rule is, that " the intensity diminishes in the same proportion as the " sine of the angle of obliquity of such surface to the direction of the rays is diminished." Again, this inclination of the earth's axis to the plane of the •ecliptic, as is evident from Fig. 2, causes the relative lengths of day and night to vary, so that, in addition to the rays being weaker in winter, the sun does not shine each day for so long a time as in summer. In short, the earth is neither heated so much nor so long, and has a longer time to get cool by radia- tion in winter than in summer. Again, it is self-obvious that when the sun's rays shine, as shown in Fig. 4, they are more refracted and pass through denser vapours, thus rendering them additionally weaker than ■when they shine as in Fig. 3. It is needless to discuss the interesting climatic and other Fig. 7.— Points on the horizon of Fig. 8.— Points on the horizon of the sun s rising and setting on the sun's rising and setting on the longest day. the shortest day. changes which may be effected by placing an imaginary specta- tor at different spots on our earth's surface. For practical pur- THE SUN'S RAYS. I J poses it will be sufficient to remember that, say in London on the longest day, June 21, the sun rises about 50° (50° 15' 20") east of north, attains a height of 62° above the horizon, and sets. about 50° (50° 15' 20") west of north, remaining above the horizon about 1 7 hours ; and that on the shortest day, Decem- ber 21, it rises about 50° (50° 15' 20") east of south, attains a height of about 15° above the horizon, and sets about 50° (50° 15' 20") west of south, remaining above the horizon about 7 hours. (See Figs. 7 and 8.) In this way the problem of the seasons will, it is hoped, b& rendered sufficiently clear for practical purposes. THE SUN'S RAYS. PLANT LIFE. As the life of plants, under glass as well as in the open air, so- essentially connects itself with the sun's rays, a glance at the scientific properties and treatment of those rays, even at the risk of stating some facts patent to everyone, will be found advantageous. Vegetable carbon owes its origin to the atmospheric carbonic; acid. By the chemical action of the sun's rays, the carbon is set free from the oxygen. The carbon assimilates the elements, of water, forming cellulose, woody fibre, &c. The oxygen returns to the atmosphere in the gaseous form. Combustion on the The term "latitude" means the angular distance of any place on the globe, north or south of the equator (in the case of England any latitude indicated would of course be north). Under the articles, " .Sun's Rays,"' "Aspect," "Inclination of Roofs," some of the few astronomical facts, mentioned here are again touched upon for the purpose of more clearly explaining the points raised. 14 THE SUN'S RAYS. Other hand takes oxygen from the air and gives to the air carbonic acid. The solar rays may be said to possess three distinct powers — lighting, heating, and producing chemical action. So distinct are thesd powers, that we can separate the heat rays from the light rays, and partially or wholly stop either at will. The rays producing chemical action can also be examined, in a great measure, apart from the heat or light rays. LIGHT RAYS. A ray of light may be partially or wholly stopped in several ways. It may be turned from its path by reflection or by refraction. It may be swallowed up, so to speak, by absorption. Or its obstruction may ensue from a combination of these causes. Light, passing through a vacuum, is transmitted without any loss ; but whenever it traverses any medium, however transparent, some of it is stopped. In fact a certain depth might be assigned to the atmosphere, which would completely intercept the solar light. About 7 feet of water is sufficient to intercept half the light of a ray passing through it. According to Dr. Letheby, the loss of gas light in its transmission through gas globes is as follows : — Table I.— Loss of Light by Transmission through Glass. Gas Globe of Clear Glass j> ») Ground ,, " )» Opal „ , Loss 12 per Cent. )> 40 >> it 60 J) But if light only reached us by the direct rays of the sun, then at noon-day an opaque body, interposed between our eyes LIGHT RAYS. IS and the sun, would simply have the effect of giving us an instantaneous midnight. That this is not so, and that during the daytime we see light everywhere, although we may not absolutely see the sun, is due to the fact that light is reflected or diffused, from all bodies. Table II. — Loss of Light by Reflection from Glass. Angle made by Sun's Rays with Glass. Rays Impinging, Rays Lost by Reflection. 5° 10° 15° 20° 30° 40° 50° 60° 70° to 90° 1,000 1,000 1,000 1,000 1,000 1,000 1,000 1,000 1,000 543 412 300 222 112 57 34 27 25 A very important fact regarding direct rays of light, which has a bearing on the subject of green-houses, is that the pro- portion of light rays lost by reflection from such a surface as a smooth plate of transparent glass, varies with the angle at which the rays impinge upon the medium. This is strikingly shown in Table II.,(p. 15), given by Bouguer. If, therefore, we wish to utilize the greatest proportion of the direct solar light rays, when transmitting them through glass, it will be obviously our most advantageous course to arrange i6 THE sun's rays. that the rays shall strike such glass at as nearly a right angle as possible. (See " Inclination of Roofs.") In consequence of the great distance of the earth from the sun, the solar rays may practically be regarded as parallel with each other when they strike the earth. When the luminous point is near at hand, the rays must necessarily be considered divergent. The law of divergent rays is that their intensity decreases inversely as the square of the distance from the luminous point. This is very clearly shown by Fig. 9, a re- production of an illustration in that excellent little book, " Philosophy in Sport." Fig. 9.— Diagram illustrating law of radiant light and heat. If a screen, say i square foot in area, be held at a yard distant from a candle, the area shaded by those rays at 2 yards would not be 2 but 4 square feet ; at 3 yards, not 3 but 9 square feet ; at 4 yards, not 4 but 16 square feet. In other words, the same rays which, at i yard distant, would give a certain intensity of illumination, at 2 yards would give, not i but I the intensity; at 3 yards, not j but i the intensity; at 4 yards, not I but ^ the intensity ; and so on. HEAT RAYS. The solar heat rays are subject to the same phenomena of partial or entire interception by reflection, refraction, and absorp- HEAT RAYS. 1 7 tion as the light rays, but the same substances have not the same eifect on both. This is exemplified by interposing a plate of glass between a bright fire and the face. The heat rays are perceptibly intercepted, the light rays are not. Solar heat rays, like light rays, are transmitted through a vacuum without loss. Through an elastic fluid such as the atmosphere, they are trans- mitted with very little loss, but when that distance is greatly in- creased, interception is much facilitated. This affords one good reason why the morning or evening rays are much less powerful than the noonday rays. But heat is transmitted in other ways than by radiation — viz., by conduction and convection. When two substances are placed in contact, one hotter than the other, the former conducts or transmits to the latter a portion of its heat till both are in equilibrium. Treated in this way, quiescent dry atmospheric air is one of the worst conductors of heat. By convection is understood the process of conveying heat by cur- rents. Convection takes place when heat is applied beneath a vessel of water, and currents conduct the heat from one part to another. Thus, altogether quiescent air may be a bad conductor, set it in motion, and it becomes a good transmitter of heat by convection. The law given in the previous section, that the intensity of divergent rays decreases inversely as the square of the dis- tance from the luminous point, applies to heat as well as light rays. In order to rightly understand the role of heat in the growth of plants, it is important to know what part of the heat rays which strike the leaves is absorbed by them, what part is thrown back and scattered, and what part passes through them to lower organs. An inquiry of this nature was made some time ago by M. Ma- quenne. Of his method we will merely say that he used, as constant heat-source, a Bourbouze lamp (in which a platinum wire is kept glowing by a regulated mixture of coal gas and air) ; B 1 8 THE sun's rays. and for some experiments with low temperatures he employed Leslie's cubes. The results of the research are briefly as fol- lows :— I. All leaves scatter a part of the heat they receive ver- tically to their surface ; with the Bourbouze lamp this diffusion is about o'zs of the whole heat, with a Leslie cube a small per- centage. 2. Generally the under side scatters more than the upper, but the reverse sometimes occurs. 3. Leaves absorb a good deal of the heat from the Bourbouze lamp, the absorption bemg due to the presence of absorbing substances, especially chlorophyll and water, in the tissue, and to the diffusion taking place internally at the surface of each cell ; it is generally greater at the upper side than at the lower. 4. Thick leaves absorb more than thin leaves. 5. The absorptive power of leaves for the heat of boiling water is very nearly equal to that of lamp- black. 6. Leaves let heat pass through better the thinner or younger they are. 7. The radiating power of leaves with a great excess of temperature is pretty near that of lamp-black ; it de- creases a little when the inclination increases. 8. The absorp- tive power of chlorophyll is, on an average, equal to. that of water for rays of the Bourbouze lamp, and increases proportion- ately to withdrawal, in one direction or the other, from the heat maxi7nuin. Additional data and remarks more nearly connected with artificial heat and its application are given in the article on heating. CHEMICAL RAYS. Besides the rays of light and heat, we have the actinic (chemically active or photographic) rays. As will be seen by the annexed table of the results of experiments by Drs. Roscoe and Thorpe, the intensity of these actinic or chemical rays varies with the altitude of the sun. CHEMICAL RAYS. Table III. — Acti.ng Intensity of Daylight. 19 Mean Altitude of Sun. Chemical Intensity. ' Sun. Sky. Total. 9 SI 38 38 19 14 23 63 86 31 14 S2 100 152 42 13 100 IIS 215 S3 9 136 126 262 61 8 I9S ■ 132 327 64 14 221 138 359 These experiments were made at Lisbon, where the sun attains a higher altitude than in England. It will thus be seen from this table, and from the preceding remarks, that at altitudes be- low 10° above the horizon, direct sunlight is robbed of all its chemical rays, and that the nearer the sun sinks towards the horizon the less intense are its heat, light, and actinic rays. From numerous simultaneous experiments carried out by the same gentlemen, it was ascertained that the average of the chemical intensity of daylight at Para was 303-2, Lisbon no, and Kew 46*06 ; or more than 6J^ times greater at Para than Kew. These figures are of value, as representing, for botanical purposes, the relative chemical activity of daylight in England and the Tropics. Although we can examine the actinic rays almost apart from the light and heat rays ; can tell, by experiment, when they are strongest, when weakest ; and can even determine that there are at least two descriptions of actinic rays (those which can com- 20 INCLINATION OF ROOFS. mence but cannot continue chemical action, and those which can continue when started but cannot commence chemical action) ; yet our knowledge of the bearing of these rays on plant life is not yet sufificiently advanced to enable us to benefit ma- terially in constructing our glass-houses. The day may come when we shall as easily regulate our actinic rays as we do our light and heat rays. From this cursory glance at the sun's rays, we see that in constructing and using our green- houses, it is of importance that we have, not only a clear idea of the properties of those rays, but possess appliances, with the knowledge how to use them, for ensuring the transmission of the maximum amount of solar light and heat rays when necessary ; of moderating their intensity when we wish ; of intercepting either one or other as we may desire ; of stopping all direct rays and utilizing only diffused rays when advisable ; of producing, directing, and con- trolling an artificial supply ; in point of fact, of turning on, adjusting, or shutting off our light and heat in accordance with our demands, with the same ease that we deal with the hot and cold water supply to our baths. INCLINATION OF ROOFS. The proper inclination or pitch of the roofs of horticultural buildings is a very important point, and depends upon a com- bination of conditions. One very necessary condition is to insure the transmission of the sun's rays in the most complete manner, and at the time when they are most required. We will first of all presume that every other condition is subordinate to this, and that we only have to determine the ANGULAR MEASUREMENTS. 21 inclination which is the best adapted for growing, ripening, and fruiting purposes, so far as relates to the sun. ANGULAR MEASUREMENTS. The pitch of a roof is the angle which is formed between the roof and a horizontal line drawn not lower than the eaves or the lowest point of the roof (see Fig. lo). If A C be the total height of vertical front, and B D the back wall, we draw a horizontal line through A. We then place a protractor on this horizontal line with the centre at the point A, when, if we require a roof of an inclination of 30° pitch, we draw the line A E intersecting the protractor at the point marked 30°. Or, if we require a roof of 40° pitch, we draw the line A F inter- secting the protractor at the point marked 40°. It must be remembered that the horizontal line through A is the base line from which all the angles of roofs are taken, as if the back wall B D be taken as a base line, mistakes will of course occur. Any pitch of roof can thus be readily ascertained. 40" 30" riG. 10. Method of determining angular measurement of roofs. With a given height of back wall, and a given width of house, the pitch of the roof will, of course, vary. Table V. 22 INCLINATION OF ROOFS. will give the varying heights of back walls above vertical fronts. The spaces are left blank where it is considered that the pitch is too low for practical purposes. By vertical front is, of course, meant the whole vertical height (generally composed of part brickwork and part glass work) before the "spring" of roof is reached. The following table may be useful as showing the angle, which is equal to a rise of 6" to 15" in the foot : — Table IV. — Angular Measurement of Roofs. Rise In Inches per Foot. fi 6 ti 7 8 9 II 10 Angle of Roof. t 26 33 / 30 15 / 33 41 » 36 52 ' 39 48 Rise in Inches per Foot. II 12 13 II 14 1/ IS Angle of Roof. ' 43 10 ' 45 47 57 ' 49 24 ' 51 20 ANGULAR MEASUREMENTS. 23 O M3 00-^ C>\-n ^ U) W •-• O O OO"^ O ^ i-ft U) 4s* .f*. to to to to tOO^C>JOJ-^-^ 00 to -VI OJ O OO"-" to On O to IsJ to to to t0OJOJOJO04^Ji».-^ HI tOOOi^ OnOOO tO»-n*^ M^^VO p-H 00 t-« 4^ 00".^ CO O to i-H O -e- W to M (O to to tOOJOJOJUJ+>'-l^4i'V^ •-H to Oi i^ Q\ OO^O M OJ On 00 M Ln OOL*J _ _ _ _, ^ OJ 4:*. OoO"^ tooj-P>.-p. 0\>-t 4^ Ui (-n i-i Oo to vO 00 O VO 00 tJ to to to t0UJOJLoOJ(-O-P».-l^4^'-n'-n 4^ *^ 0\^ \0 O W -^ 0\0 M Wi 00 to ON t-«tOOJ«-^tO\Ji4:k4s*Vai-i«j-i to "-I 4i. i-i Ul -1^ '-' 00-^ to to-^so O to-^oo to to t0OjO0OiUiOJOJ-f^-P^4^'-rnjT-a ONW \0 O t0O0'-n'vj\O tOLn OOMt.nvD Oj4i^ to 4:iOJOJ-l^t-« to O <^OJ OOO '-H\04^000\0 O O O W t0OJC^OJO0O0U)-ti»-4^4i^-(^'-"'-"'-" ON 000 •-' t04i. ONOOOOJ'-nVJ OOJVJ i-t 4^ to LfT l-( HH »-i \0 -^ ON-P*- O '-" "O 4^ O O 4^ 4^ (^n C^ to 4^ W CO to 4^ OJO0U>OJOJOJ^^4^4i-"-n'-"'-"»-n ON O tOUit-n OnCOO tOi-n OsOUJ ON^O UJ 1./TW 1-1^^4^.00 uit-i i-(4i».tO 00 ONt-n OJ to O ONOJ COW OOVO Vx "^^ OJ OJ 00 00 U) 4^ 4^ - U»4i>.v^'^vO O W^ to Ui, to V/H-n 00^ O 4^ *-"'-" ^>> k ^^ "»n Ln i-n On On I VO to WT 00 "-I WT T 4i. to i-H to 44. »-t J ON ON 004^ to OJ c_io OJ 00 00 4^ 4^ 4»- un ON -^vO t-H 00 "-ri to UT to »-< O 00 to 00 fH O O 4^4^ to wi ^J^ Ln On On On »-< 4!k 'vj o U» ON v^ to 1-1 )-H to 4^ i-i ^0 ON ON ON CO 00 OJ 00 4^ 4^ -^ 4^ - ON OO^O i-i 00 WT ^ V I ui '^n i-n On On On 1 OJ ON^ i-i »j^ 00 \\^n M rT\ n m HH fll w B H > z D o •z INCLINATION OF ROOFS. BEST TRANSMISSION OF SOLAR RAYS. From Bouguer's table (see the "Sun's Rays,") it will be seen that the more nearly do the solar rays form a right angle with glass, the less is light lost in transmission. It may, there- fore, be necessary to construct a roof of such a pitch that the maximum amount of sun's ray's shall be admitted at or about a certain date, when all the sun power which can be obtained is required for ripening or other purposes. Thus, supposing A B (Fig. ii) to represent a sheet of glass, taking the total rays of light impinging as looo : Fig. II. — Loss of solar rays in transmission through glass. The numbers near the outer semi-circle will show the rays lost by reflection, and those near the inner semi-circle, the angle of incidence. From this diagram we see that if the sun's rays strike the glass at any angle between 60° and 90°, only about 25 out of 1000, or 2j^ per cent, of the sun's rays are lost. This allows us a range of 30° on each side of the perpendi- cular, with the minimum loss of rays. Now, if a glass roof be inclined so as to face the south at an angle equal to the latitude of the place, a greater amount of solar rays will pass through such a roof at, say the equinoxes, than if it were in- clined at any other angle. But according to the above diagram, we may safely deviate at any rate, 20° on either side, so that if, when built at an angle of BEST TRANSMISSION OF SOLAR RAYS. 25 51°, a roof receives the sun at an angle of 90°, we may incline our roof at any angle between 31° and 71° without materially affecting the amount of rays transmitted through it. This allows us a good margin for the consideration of other points in the determination of the pitch of our roofs. If, however, we desire accuracy of the highest degree, and require that our glass roof shall, on any particular day in the year, receive the rays of the sun (when at its highest altitude) at a perfect right angle, we have only to take the latitude of the place, and from it subtract the sun's declination between the vernal and autumnal equinox, or to it add the sun's declination between the autumnal and vernal equinox, and we shall obtain the angle of inclination of the roof. This will be self-evident from the annexed figure : — Fig. 12. — Relation between inclination of roof and latitude. A B C D represents the section of a green-house built upon the earth at latitude 51° north. The roof A B is constructed with an inclination of 51° to the base line. The arrows show 26 INCLINATION OF ROOFS. the rays of the sun striking the roof A B at right angles. Th« figure is drawn supposing the equator F G is coincident witb the plane of the ecliptic. As the two planes intersect one another on March 21 and September 23, the figure would be correct for these two dates only (that is, at the two equinoxes), Suppose, however, we wish to find in the same latitude, 51° the angle of elevation which a roof should have to receive, or say August isth, the sun's rays at right angles to the glass, 51°, the latitude, minus 14°, the sun's declination on Augusi iSth, equals 37°, the angle of elevation of roof. Similarly, there is a variation of 1° in the pitch of roof for every corresponding variation in the latitude, so that, adhering to the same day August ISth, Table VI. will give us the inclination of roof fo) any latitude between the north of Scotland and the south o England. Table VI. — Pitch of Roofs at Various Latitudes. Latitude. 50° 51° 520 53° 54° Most favourable pitch of Roof. 36° 37° 38° 39° 40° Latitude. 55° §60 57° S8° 59° Most favourable pitch of Roof. 41° 42° 43° 44° 45° But August 15th is nearly eight weeks after the longest daj so that the sun's rays will also fall perpendicularly to the roc about April 27 th, or eight weeks before the longest day. Wit this arrangement the sun's rays would only deviate about ic at Midsumraei-, a deviation which is, as explained previousl; immaterial. The immediately preceding table may, therefore VARIOUS CONDITIONS. 27 be taken as indicating the most favourable pitch of roofs in various latitudes which can possibly be constructed. VARIOUS CONDITIONS. The foregoing remarks are intended to apply more parti- cularly where but little artificial heat is used, or where sun-heat is required under the most favourable conditions. Where ample fire-heat is available, the pitch of roof may be allowed to depend more upon other conditions. These may be sum- marised as follows : — Fixed height for back wall. Fixed height for vertical front. Fixed width of house. Inclination of roof which shall suit construction best. Influence of inclination of roof on heating. Distance of roof from plants. Inclination of roof in relation to condensation and drip. If the height of an existing back wall cannot be altered, and the width of the house be determined by some local con- dition, we can yet vary the pitch of roof by raising or lowering the vertical front. Sometimes it is necessary to adhere to a definite height of front, to a definite width of house, and to preserve a back wall which is not high enough for the purpose of a desired inclination of roof. It will frequently be possible to overcome such a difficulty by giving the roof the required slope until a certain point is reached, where, by dropping the roof at the same angle, the top of the low back wall is just reached (see Fig. 29). Of course, where it is necessary to have a continuous rafter, undisturbed by the break of a ridge, the space between the top of the roof and the top of the wall must be filled up by a vertical glass spandril. A roof must not have an inclination which is entirely un- 28 INCLINATION OF ROOFS. suited to constructional requirements. For instance, it i found by experience that if a glass roof have a lower pitcl than about 26°, or 6" risfe in a foot, rain is apt to drift unde the laps. Sometimes, especially in span-roof conservatories, 1 pitch of 30° is convenient, as a circular strengthening rib ti rafters can be more easily introduced in a roof having 30° pitcl than any other. (See " Show-houses.") As will be seen elsewhere (see article on " Heating "), it i important that all the air in the interior of a hot-house shouL be heated as equally as possible, for the sap has a natura tendency, not only to flow to the highest part of plants, but ti flow to the spot where the greatest heat exists, so that th higher the pitch of the roof, the greater the tendency for ho air to accumulate near the top of the house, and for the sap ti rise to the higher parts of the vines, plants, &c., to the detri ment of the lower parts. The lower the pitch of rool consistent with the due admission of solar rays, the less pro bability of a difi'erence of temperature between the upper an( lower part of roof. It is also desirable that the plants should not be grown toi far from the surface of the glass, and due regard must be ha( to this point in determining the pitch of roof. Again, the lower the pitch, the greater will be the likelihood moisture condensing and dripping on the leaves. The highe the pitch, the longer the length of rafter available for vines &c. The question of roofs in relation to drainage is discussei in another place (see "Drainage"), but it may be remark© that the weight of snow on a roof varies from 3 to 10 lbs. pe foot super, according to climate. The pitch of roof has also an influence upon constructiot for if a roof be not properly " tied," a high or low pitch ma greatly affect lateral thrust. When one roof meets another at right angles, it is advai ASPECT AND SITE. 29 tageous for both roofs to have the same inclination, otherwise the roof-bars will not properly intersect, and the line of juncture will have a very awkward appearance. The pitch of roof may be required to deviate from an other- wise suitable inclination where wall-fruit only have to be preserved. (See "Wall-tree Protectors.") It may be taken for granted, that where great heat is necessary the plants require to be near the glass, and houses are built low, 26°, or 6" in the foot, is a very good inclination of roof Where the solar rays are important for ripening purposes, from 36° to 44°, in accordance with Table VI., p. 26. For general purposes, and as may be determined by other conditions, from 26° to 44°. And for a specially narrow, high form of house, such as a wall-tree cover, a greater pitch than this is required. Roofs with more than one pitch, and curvilinear roofs, are treated in other sections. ASPECT AND SITE. Horticultural buildings usually require to be so placed that they may receive the maximum amount of the sun's rays. The position of the sun (see " Astronomical ") when it reaches every day its maximum height is somewhere between the southern point of the horizon and the zenith; in fact, between altitudes 15° and 62°. Consequently, if a wall require to have a house placed against it, and this house be of the lean-to de- scription, other circumstances being suitable, the wall should run east and west, and the house should be built on the side of it facing south. During that part of the year when the sun rises north of the east, and sets north of the west, such a house would, of course, lose part of both early morning and late afternoon sun. To obviate this objection, if the wall be not 30 ASPECT AND SITE. too high, a three-quarter span-house can be erected. Any buildings, out-houses, &c., on the north side of such a wall would afford additional protection from cold winds. If it be necessary to build in the open, a span-house is usually considered the best. In such a case the ridge may run north and south, and the sides face east and west. In this manner each side receives its due share of the sun's rays,- the side facing east receiving the whole of the morning and part of the afternobn sun, the side facing the west part of the morning and the whole of the afternoon sun. Sometimes it is considered advisable to reverse this aspect for a span-house, by letting the ridge run east and west, and the sides face north and south. In this case the preponderance of sun's rays will, of course, be re- ceived on the south side, leaving the north side much cooler. Plants requiring different amounts of solar rays can therefore be grown in the same house. But it sometimes happens that the choice of aspect is not perfectly open. Perchance it is necessary, in the case of a lean-to, to place the house against a wall which does not face the absolute south point, but is inclined to the east or west. It must be remembered that in such proportion as the east end approaches the south, by so much is the early morning sun lost ; and in such proportion as the east end approaches the north, by so much is the afternoon sun lost. If the deflection of the wall from the line of east and west is great, one means of overcoming the difficulty is by building a three-quarter span- house, the rays which would otherwise be lost on the back of the wall would then be caught. The deflection of the wall from the line of east and west may be so great, that the wall may absolutely be north and south. Several courses are then open. The wall may be made sufficiently low for a span-house to be built. Or a three-quarter span-house may be erected Or if the wall be too high for such a house and a lean-to must ASPECT AND SITE. 3 1 be erected against the wall, the greater proportion of the morn- ing sun will be lost if the house be on the west side, and the greater proportion of the afternoon sun will be lost if the house be on the east side of the wall. Or another site altogether may be adopted. If a combination of houses require to be erected, the broad principles of the foregoing remarks will hold good. In the case of a quadrangular walled garden, with the walls facing respec- tively east, west, north and south, there is no objection, so far as aspect alone is concerned, for lean-to, or three-quarter span-houses, to be placed on all the four walls. The amount of the sun's rays or the time during which they will have in- fluence upon the houses, will thus be varied in four different ways. The house on the south wall will have the maximum amount for the longest time. The house on the west wall will lose a great proportion of the morning sun, but will catch all the afternoon sun. Such an aspect is frequently considered advantageous, in order, where no (or very little) fire heat is available, that the latest sun heat may be retained by closing the ventilators and thus counteracting, as much as possible, the fall of temperature during the night. The house on the east wall will receive the same amount of solar influence as that on the west wall, but in the morning instead of the after- noon. The house on the north wall will have the smallest amount of solar influence, but in many cases a specially cool house is desirable. In all remarks a west (&c.) house or wall means a house or wall facing west (&c.). In deciding upon the question of aspect, much will depend upon the description of plants, &c., grown, the amount of ground at disposal, the dimensions of the horticultural buildings, economy of space, compactness of the structures, accessibility and various other local and incidental conditions. In selecting the site, care must be taken in observing whether 32 LEVELLING. trees or other objects are likely to obstruct the sun's rays. The rights of neighbours must not be trespassed upon, and the regulations of local and other authorities must be observed (see Legal). The question of levels and drainage, &c., must also be very carefully studied. If houses are to be erected, in which, although they may be structures for purely growing purposes, the ladies of the family take an interest and may be expected to visit occasionally, their comfort and convenience should be studied in the selection of a site. LEVELLING. This is such an important point in relation to erecting horticul- tural buildings, draining, and laying out gardens, that a few words are absolutely necessary on the apparatus required, the elementary principles of levelling, and their application to houses. APPARATUS REQUIRED. A gardener will find it useful to have — Bricklayer's ordinary Fig. 13.— Levelling instrument. ELEMENTARY PRINCIPLES. i3 horizontal and vertical spirit and plumbob levels : a reel and line ; stakes ; measuring rods ; an ordinary Gunter's chain, which is 66 feet long, having in it loo links (it will be re- membered there are 1 00,000 square links in a square acre) ; a drainage level (as shown by Fig. 13) ; and two staves. (See also tools, &c., mentioned in the section " Tool-house.") ELEMENTARY PRINCIPLES OF LEVELLING The earth is spherical, and its surface forms a circle, having a radius of 3,956}^ miles; consequently an apparent level taken on the earth's surface would form a straight line, whilst the true level should, of course, form part of the above circle. Table VII. shows the height of the apparent above the true level from 100 to 1,000 yards : — Table VII. — Difference between Apparent and True Levels. Yards. Inches. Yards. Inches. 100 0-026 1 600 0-928 200 O-I03 700 1-264 300 0-232 800 1-651 400 0-413 900 2-089 500 0-645 1,000 2-584 Correction of the levels is also required for the refraction of the atmosphere. Although this varies, it may generally be taken at t of the correction for curvature. Refraction makes objects appear higher than they really are, therefore the i must be deducted from the correction for curvature. c 34 LEVELLING. Although it is necessary to mention these points, the errors due to curvature and refraction may be avoided by placing the levelling instrument in the centre of the particular section of ground which is being levelled. Suppose the difference of true level between A and E (Fig. 14) requires to be found, the whole distance must first be split up into convenient levelling sections, of which, in the present instance, we will imagine there are UWO— A C and C E :— Fig. 14. — Method of long distance levelling. Staves are fixed at points A and C, the instrument placed midway between them, and the readings of the two staves noted. The operation must then be repeated between C and E. The difference between the sum of the backsights on A and C, and the sum of the foresights on C and E, will show the difference of level between A and E. Thus we will suppose the readings between A and E to be as follows : — Feet. Feet. Backsight from B on A 10 '44 Foresight from B on C 5 '63 Backsight from D on C 9*53 Foresight from D on E 3*58 19-97 9"3i 9'3r Total rise from A to E io'67 LEVELS FOR HORTICULTURAL BUILDINGS. 35 When the backsights exceed the foresights, the ground rises j when the foresights exceed the backsights, the ground sinks. If the total distance to level be great, and intermediate points cannot conveniently be taken, correction for curvature and refraction must be made. This, however, is not likely to occur in connection with horticultural buildings or a garden, unless the exact difference between the levels of a certain spot (say a delivery tank and the distant source of water supply) be required. When the ground to level is of small extent, the following method may be adopted : — Drive pegs into the ground, in the direction the level requires to be taken, at rather shorter intervals than the length of the ordinary bricklayer's horizontal level. (See Fig. 15) : — Fig. ij. — Method of short distance levelling. With peg No. i as a fixed point, raise or lower peg No. 2 till the bubble in the spirit tube of the level be central. Then, with No. 2 as a fixed point, raise or lower No. 3 till it be level with No. 2. Then take No. 3 as a fixed point, and repeat the operation with No. 4, and so on. A length of ground up to 100 or 150 feet may be very accurately and quickly levelled in this manner. LEVELS IN REGARD TO HORTICULTURAL BUILDINGS. An important object in selecting a site for horticultural buildings, on ground which is undulating or has a general slope, is that there shall, if possible, be rising ground or protection on 36 LEVELLING. the north, and a fall, or, at any rate, no obstruction, on the south. Another object is that the buildings shall be capable of easy drainage. When the ground is fairly level in the direction of the length of the buildings, but rises or falls in other directions, artificial levelling of the ground is seldom required. When, however, the ground is not level in the direction in which the houses are built, several courses are open. First, the house may be made to retain its floor line at the highest level of the ground, as at A in Fig. i6 : — Fig. i6. — Longitudinal block elevation of a house embanked. Secondly, if the house or houses be very long, and surplus soil can neither be obtained or disposed of, other things being equal, the level may be obtaiiled by half excavating and half embanking, as in Fig. 17 : — b^ Fig. 17. -Longitudinal block elevation of a house half excavated, half embanked. Thirdly, if by reason of the excessive fall of ground or length of houses this is not practicable, step-levels may be introduced, as in Fig. 18 : — Fig. 18.— Longitudinal block elevation of house on step-levels. LEVELS FOR HORTICULTURAL BUILDINGS. 3/ In such a case the boiler would have to be at the lowest end, or at such a depth and in such a position that it can supply, with ease, the lowest house. This third course should be avoided, if possible, as it is not generally advantageous to have the paths in a range at different levels. (See " Paths.") In case, from any reason (such as a difficulty in drainage) the general level of growing-houses has to be raised, it is frequently advisable to reach such level by an inclined plane rather than by steps, so that a barrow may be wheeled direct into the houses. It is also advantageous to arrange' that the floor level shall coincide with the ground level at one end of a greenhouse or range of houses, in order to ensure ease of access for a barrow. On no account whatever should the line of glass and wood-work be allowed to follow the slope of ground either as in Fig. 19 or 20 : — Fig. 19. — Longitudinal block eleva- FiG. 20. — Longitudinal block eleva tion of an unlevelled house tion of an unlevelled house (incorrect). (faulty). The author has seen such houses, but for ordinary horticul- tural work they are unnatural, inconvenient and hideous. Sometimes long corridors are constructed as shown by Fig. 20, but they would be much more sightly, if, while the floor-line followed the slope of the ground, the roof were to be con- structed of step-levels, as shown by Fig. 18. A good instance of this is the long corridor leading from the Crystal Palace to the London and Brighton Railway Station. All the above remarks will apply to combinations as well as single houses. In the case of combinations, however, the 38 DRAINAGE. houses composing each set should, if possible, be on the same level. Parallel lines of houses may occupy different levels. In fact, while not sacrificing cost, sightliness, utility, &c., to uniformity in levels, aim at the latter whenever possible. The floor-lines of conservatories and horto-architectural structures, when not dependent upon the level of other reception-rooms, should be two or three steps above ground-line, as much to produce effect as to facilitate drainage. The floor-level of a conservatory which communicates with a dwelling-house should be about the thickness of a mat, but no more, below the level of the communicating reception-rooms. (See "Paths" and " Show-houses.") DRAINAGE. GENERAL REMARKS. It would be possible to give a large mass of information on the subject of draining land. Such information would embrace geological data, with deductions, and a branch of mechanics, which, however useful to the gardener, would be irrelevant to the present work. It is therefore proposed to give merely one or two useful tables, a few general remarks, and then some practical notes on drainage in relation to horticultural build- ings proper. In planning drains, Hurst recommends that the figures given in Table VIII. be taken to represent the quantity of water per hour for which draining provision should be made. GENERAL REMARKS. 39 Table VIII. — Provision for Draining Various Surfaces per Hour. Draining Surface. Ins. of Water in depth. Gallons per i,aoo square feet. From Roofs (horizontal measure) ... •s 260 „ Paved Surfaces •I 52 „ Gravel (Clay Subsoil) •05 26 ■ ,, „ (Gravel Subsoil) •01 S ,, Meadows or Lawns •02 10 In open country, where the soil is loose and permeable, about one-third only of the rainfall finds its way into the water-courses, the remaining two-thirds being absorbed or evaporated. Table IX. will give other data in regard to the drainage of land. Table IX.— Drainage of Land (Molesworth). Soil. Depth of Pfpes. Distance of Pipes apart. ft. in. feet. Stiff Clay 2 6 IS Friable Clay 2 18 Soft Clay 2 9 2t Loam, with Clay 3 2 21 ,, Gravel 3 3 27 Light Loam 3 6 33 .Sandy 3 9 40 Light Sand, with Gravel ... 4 5° Coarse gravelly Sand 4 6 60 40 DRAINAGE The nature of the surface of the soil must be taken in con- junction with the subsoil and direction of the strata. A garden may slope towards the south, and yet a retentive undrained ubsoil may neutralize surface advantages. Any chilling effect vrhich spring water in the soil may produce must be guarded against. Open drains and covered rubble drains are used in land drainage, but pipe drains are chiefly adopted. The angle on plan, which a subsidiary drain makes with a main drain, is frequently 90°, but before making the juncture, the former should gently curve so that, on joining, its contents may flow in the same direction as the contents of the latter. The former should, moreover, whenever possible, have a decided fall into the latter. In draining a garden, especially of clay or heavy soil, it is frequently advantageous to sink a central well, with a dome- top rising just above the ground, and having a man-hole in the centre covered by a substantial trap-door. Drains from the vari- ous parts of the garden may be discharged into this well, which should be provided with an overflow pipe at a lower level than the inlet discharge pipes. By opening the trap-door, it can easily be seen if all the drains are working properly. The dis- covery of a stoppage in any one of the drains- is thus facilitated. DRAINAGE OF HORTICULTURAL BUILDINGS. If the ground descend in any direction towards the build- ings, care should be taken, especially if the subsoil be non- absorbent, that due provision is made for surface as well as underground drainage, and that the water does not find an artificial reservoir in the stoke-hole, main pipe, trenches, borders, &c. If there be no other means of keeping the water out of any low lying part, such as a stoke-hole, either GROWING-HOUSES. 4I sink" a galvanized iron tank in the ground and build the stoke- hole in it (in this manner keeping the water out instead of in) or else sink a well and.drain into it. It is perfectly useless to think of keeping refractory water out of a stoke-hole by build- ing the latter of concrete, hard-faced bricks, hydraulic cement, or any combination of the three. Of course there is one other course open when a stoke-hole cannot easily be drained, namely, to raise the datum floor-line of the houses above the level of the surrounding ground, when the stoke-hole need not be sunk so low as would otherwise be necessary. When deciding the question of drainage, it is advisable to make a note of the most suitable positions for the down rain water pipes from the roofs. Down pipes should never be con- nected with the drains, but should discharge into gratings. In this manner the down pipes can easily be kept clear of obstruc- tions. When constructing drains for horticultural buildings it is often necessary to unite to them the drains for carrying away Surplus water from stables, outhouses, &c. The question of the drainage of horticultural buildings has such an intimate connection with water supply, that the foregoing remarks should be read in conjunction with the sections upon the collection, retention, and distribution of rain water. GROW I NG-HOUSES. LEAN-TO HOUSES. The raison d'etre of this form, as indeed of the other forms of glass-houses, has already been shown. (See "Aspect and Site," p. 29.) Lean-tos are generally used : — When a wall or 42 GROWING-HOUSES. building already exists, against which it is desired to place a glass-house — When a wall is specially built, in order that a house or combination of houses may face the south and have brick protection from the north — When the exigencies of the plants, &c., to be grown, demand this form. Lean-to houses facing south are easier to heat than span-houses, and are, other things being equal, cheaper to build than any other form. It must not be forgotten that even when facing exactly south, lean-to houses lose, in the hottest weather, the early morning and late evening sun. Also that a plant, if unmoved, does not catch the sun's rays all round it in a lean-to. The most usual form of lean-to is shown by Fig. 21, below. This house has about 2' 6" front wall and 2' 6" vertical light, making about 5' o" in front. These, varied a few inches either way, make very good front dimensions of lean-to for most growing purposes. For plants, 2' 6" forms a very good height for front stage ; or a raised border or internal bed suffices to c^ Fig. 21.— Section of lean-to house with front lights. lessen the height for stems of vines, &c., which are required to be trained along the glass. In this form, top ventilation is shown by lights hinged to ridge, the bottom ventilation by front LEAN-TO HOUSES. 43 lights swinging from the gutter plate. Of course, when con- sidered advisable, these lights may be fixed and ventilators built in the brick-work, as shown by dotted lines. Or both may be used, especially when, as in severe weather, it is some- times required to heat the incoming air before reaching the plants, by admitting it near the pipes placed immediately be- hind the front wall. Another lean-to, see Fig. 22, has the same form as shown in (P? Fig. 22. — Section of lean-to house without front lights. Fig. 2 1, but without vertical glass lights. Bottom ventilation is provided by swinging shutters in brick-work, top ventilation by sashes hinged to ridge. About 4' o" forms a very good height for front wall. This is high enough to admit a front stage for bedding stuff, &c., and low enough, with a raised border or inside bed, to accommodate vines, &c., to be trained near the glass. With equal areas, Fig. 22 can be built some- what cheaper than Fig. 2 1. By decreasing the height of the front, Fig. 22 is found advantageous for those trees, &c., which require a minimum space between the soil they grow in and the spring of roof. Of course, this is presuming that head room is not required inside near the front. Another form of lean-to is shown by Fig. 23. This has a broken roof, or roof of more than one pitch, a very 44 GROWING-HOUSES. low front wall supplied with ventilators, and top ventilators similar to the two preceding figures. The first, or steepest slope, is sometimes made of framed lights, which are capable Fig. 23. — Section of narrow lean-to house, broken roof. of being taken off when a very large amount of air requires to be admitted to the house. This form is frequently used where wall-fruit requires protection, and the house must be narrow, so that there may be the minimum distance between the glass and the trees on the back wall, thus counterbalancing any dis- advantage which may arise by the main portion of the roof being of a higher pitch than is theoretically advisable. Such a, house may sometimes be advantageously fitted with vertical front lights 2 feet high, coming down to within 6 inches of the ground, in order that strawberries, &c., may be grown immedi- ately in front of the lights. To cause as little obstruction as possible, where there is an outside as well as an inside border, part of the front wall may give place to iron pillars and slate slabs. (See " Brick-work and wall-tree protection.") Another form of lean-to house, but one which is not often seen, is similar to Fig. 22, but instead of having any vertical front, the roof springs practically from the ground line. The LEAN-TO HOUSES. 45 advantage of it is, of course, that the sun's rays come directly on the ground, and no length of stem whatever need intervene between the border or bed and spring of roof. This very fact, however, may be a disadvantage, as it is not always possible that the soil of the front part of the border or bed can be pro- tected by the shading of foliage, and it may sometimes be ad- visable to have a small dwarf wall in front to assist in shading the soil. In addition to this, although such a form of lean-to admits of top ventilation in the ordinary way, it does not easily admit of bottom ventilation. The only way such a house can be at all adequately ventilated, is by having sashes opening in a line parallel with the rafters. This mode of ventilating from the bottom to the top of a roof is not always considered advis- able, especially if the house be heated, and the temperature of the incoming air requires to be raised before touching the foliage. In all these figures an imaginary ground line is shown at the same level inside the houses as out. But of course it may be necessary that this datum line should vary, as, for instance, if the houses have to be partially sunk to afford minimum ob- struction, or have to be raised by reason of defective drainage or inability to sink stoke-hole so low as necessary. Or, for architectural purposes, the exterior level may require to be much lower than the interior floor line, as it may often happen that an important appearance must be given to the exterior, without increasing the height inside between ground and spring of roof. In this case, the interior datum line may be level with the top of dwarf wall, and the exterior ground line 2' 6" or 3 feet lower, so that a house may have 8 feet front vertical height outside, and only 5 feet front vertical height inside. It must not be forgotten, too, that, other things being equal, it is more costly to sink a house in the ground, than to have the floor at, approximately, the same level as the exterior ground. 46 GROWING-HOUSES. SPAN-ROOFED HOUSES. Next to lean-tos, span-roofed houses possess the most natural and advantageous form. They are very useful : — ^Where no wall exists or is required — For building at right angles to and in combination with a range of lean-to houses against a south virall — When the minimum height is required, so that there may be as little obstruction caused as possible — When plants require to be as near the glass as practicable — Or when the length of the houses must be in the direction of north and south, and each side requires an equal amount of the solar rays. Usually, span-houses are placed with the ridge running north and south, so that there may be as perfect a distribution of the sun's rays as possible on every side of a plant. Or, if it be advisable that one side of the house shall be hotter, and receive a greater proportion of the sun's rays than the other, then span- houses are frequently placed with the ridge running east and west, and such houses for a great many purposes are extremely useful. A general form of span-house is given in Fig. 24 : — Fig. 24. — Section of span-house, with front lights. This is shown with 2' 6" brick-work, and 2' 6" vertical light. All these vertical side lights can open as bottom ventilation, SPAN-ROOFED HOUSES. 47 with the substitution or addition of ventilators in brick-work, especially for use in severe weather, or when the external air requires to pass over pipes before coming near the plants. The top lights alternately on each side of ridge, or all the lights along one side of ridge, or all the lights on both sides of ridge, may be made to swing open, as local circumstances, the pro- duce to be grown, &c., may demand. Taking away the side lights, we have Fig. 25 : — Fig. 25. — Section of span-house, without front lights. This is practically the span counterpart of the lean-to house, Fig. 22. The top remains the same as in the previous illustra- tion, but the lower ventilation is provided by flap, hit-and-miss, or other approved ventilators in the brick walls. 4' o' height to eaves is shown, but this may be varied to suit the different purposes for which the house may be required. This dimen- sion, however, is found suitable when either staging, a raised border, or an artificial bed is required. With a uniform width and length of house, Fig. 25 would not cost so much to build as Fig. 24. Another form of span-house is given in Fig. 26. This shows 6' o" of vertical glass-work coming practically down to the ground line, ventilated at top and bottom, in the same manner as Fig. 24. This is a useful form of house : — When a fairly large cubical contents is required, in order that the tem- 48 GROWING-HOUSES. perature of the interior may not fall so rapidly— When an un- FiG. 26.— Section of span-house with high glass sides, ■heated structure is necessary, and the sun's rays must be ad-' mitted with as little obstruction of walls as possible— When large plants require to be planted in open borders. , Fig. 27.— Section of large span-house.— A A, casement lights. Fig. 27 shows a span-house of a larger and more pretentious SPAN-ROOFED HOUSES. 49 character. The lantern breaks up the roof, and while, by the additional framework, extra obstruction is presented to the sun's rays, yet for a large and important style of work such a form frequently becomes necessary. Such a house, although some- times used as an orchard-house, cannot strictly be called a growing-house. It is, by the height to eaves, and general pro portions, more suitable as a show-house, floral corridor, horti- cultural promenade, conservatory, &c. Another form of span-house is that in which the roof has more than one pilch. This does not possess the same advan- tages which are attached to certain lean-to houses having a double pitch. If used as a plant-house fitted with interior stages of a normal height, and the glass is practically carried down to the ground, all the glass below the stage-line is not only useless, but the external appearance is not so pleasing, in consequence of the stage-line cutting the view of the glass. Efficient ventilation of this form of house is not so easy. The first pitch of roof is almost certain to be too steep in accordance with the principles laid down in the article "Inclination of Roofs," p. 20, et seq. Plants near the lower part of the roof fare pretty well, but near the centre of the house would be much too far from the glass. For ordinary hot-house purposes such a form has no great advantages. For vinery purposes this house would be much more suitable, were bottom ventilation capable of being more easily carried out, say by having 2 feet dwarf walls fitted with opening ventilators. Another form is that in which the roof practically springs from the ground line, and really may be said to be the same as shown in Fig. 25, leaving out the dwarf walls and lower ventilators. The sun's rays can certainly reach every part of the internal area, but are apt to bake the ground near the sides. Top ventilation is, of course, easy, as in any other of the span roofs, but bottom ventilation is difficult ; or, in fact. 50 GROWING-HOUSES. any ventilation by which air requires to pass near hot water pipes before touching the foliage. This may be a cheap form of house, but cannot be termed efficient. Curvilinear span-houses are treated in an article specially devoted to this form of structure. In a lean-to the solar rays are directed to a plant principally on one side, and there is part of a plant which, if kept in a lean-to unmoved, receives no direct rays at all. In a span, however, the rays reach both sides of a plant, and more uniform growth is promoted. Other things being equal, heat is not so easily retained in a span as in a lean-to. This is especially the case if the ridge be in the direction of east and west, as radiation is more easily carried on by a glass roof and lights facing north than by a brick wall, especially if the latter be protected by potting-sheds, furnace-room, mushroom-house, &c., &c. Therefore, a span-house requires more heating power to produce the same' effect than a lean-to of corresponding area. The area of roof in a span-house is equal to that in a lean-to of the same pitch and width, but, of course, the difference in vertical height between the lowest and highest points of roof is twice as much in the latter as in the former. Thus it will be seen that span-houses are more advantageous for plants which require to be near the glass, and the necessity is obviated of having such a lofty stage as would be required in a lean-to. Span-houses of an ordinary width are applicable where no great length of continuous rafter is required. In all cases the same remarks as in the section on " lean-to houses " apply to spans as regards interior and exterior datum lines being adapted to suit drainage, stoke-hole depth, appear- ance, architectural requirements,^bstructIon of view or light, and cost. THREE-QUARTER SPAN-HOUSES. 51 THREE-QUARTER SPAN-HOUSES These are very frequently termed half-span, but it will be seen that " half-span " really indicates a lean-to, thus : — i±l Fig. 28. Therefore a three-quarter span is the better term for a roof of unequal span, in which one side of the span is longer than the other. The annexed figure shows one of the usual three- quarter span-houses : — Fig. 29. — Section of three-quarter span -house. They are useful : — Where a back wall requires to be moderately low — Where a maximum length of rafter is not necessary — Where, by themselves or in combination with other houses, it is advisable to let light in at the back — Where a certain in- clination of roof is required without altering the width or raising any more than is absolutely necessary, an existing back wall — Where heat requires to be saved by truncating the apex of an otherwise lean-to — Where the height of the back wall of a lean- to is objectionable, but yet brick protection is desired on one 52 .GROWING-HOUSES. side, in combination with a means of easily providing for potting-shed, tool-house, mushroom-house, &c. Three-quarter span-houses are applicable to much the same conditions as lean-tos, and the remarks in the preceding section will generally apply to this. Their best aspect is to face the south, when they catch, not only the same amount of sun's rays as lean-to houses, but what the latter fail to catch, viz., all the rays when the sun rises north of the east and sets north of the west, or the early morning and late afternoon sun during the long days. Like lean-tos, three-quarter span-houses, when facing south, are easier to heat than span-houses of correspond- ing dimensions, owing to the protection of the north wall. Supposing the pitch on each side of the ridge is the same, the area of roof is exactly equal to that of a lean-to or span of the same length, width and pitch. The three-quarter span form is very convenient in some cases when it is desirable to have the top ventilation hidden from sight, or when, for architectural effect, an ornamental ridge is advisable, yet the house requires to approximate as nearly as possible to the lean-to form. Unless in combination with other houses, or to fulfil some special object, three-quarter span- houses may be generally considered unsightly. CURVILINEAR HOUSES. These have been extolled at various times as possessing the most advantageous form for growing purposes, but upon a con- sideration of their claims, it is open to question whether their idvantages are not overwhelmed by their disadvantages. Other things being equal, a curvilinear roof doubtless admits the solar rays in a better manner than any other form of roof. .But it must be remembered that the sun's rays are, so far as the earth is concerned, practically parallel ; so that at no time a CURVILINEAR HOUSES. 53 can they impinge upon the whole of a curvilinear roof at right angles, as they can upon a straight roof, thus : — Fig. 30. — Vertical block section, showing sun's rays striking a curvilinear roof. Fig. 31. — Vertical block section, showing sun's rays striking a • flat roof. Suppose the rays impinge on the centre of the roof, Fig. 30, at 90°, it will be seen that the angle steadily decreases as the rays^approach the bottom or top of the roof. This, however, is notj^of so much importance when we consider that the angle of incidence may vary 20° or 30° without decreasing the amount of sun's rays transmitted. Then again, suppose the centre of the curve is taken at the bottom of the back wall, as in Fig. 30, the top of roof will approach the horizontal, and the bottom of roof will approach the vertical, thus increasing the liability of rain to drift in at the top, between the laps, and also form a lodgment for the snow. This may, of course, be remedied by making the centre of the curve below and behind the bottom of the back wall, thus : — • Fig. 32.— Vertical block section showing, a mode of determining the radius of curvilinear roof. This mode cuts off the extreme vertical and horizontal portions. 54 GROWING-HOUSES. But even supposing this done, it is difficult to construct venti- lators which shall form part of the curve. The construction of the framework of a curvilinear roof, either in wood or iron (in small work), is more costly and troublesome than that of a straight roof, and lateral thrust is not so easily overcome. Training wires are not so easily fixed, and heavier iron vertical rods are required if it be necessary to follow the exact curve of roof. If bent glass be used, it is not only much more costly at first, but much more expensive to keep in repair and much more susceptible, by changes of temperature, to breakage than flat glass. If, to obviate these disadvantages, the roof is made up of straight panes of glass, these must be short, or they will not follow the bend of roof. In either case they look bad, unless the curve has a very large radius indeed. The glazing of straight pieces of glass ona bent rib is also not easy. Taken as a whole, circular work may in a few exceptional instances be introduced to obtain an architectural result, or in moulding the lines of a large winter garden or magnifi- cent palm-house, but for ordinary growing purposes, we may consider curvilinear roofs not so suitable as those composed of straight lines. RIDGE AND FURROW HOUSES. By a "ridge and furrow" house is usually meant a house having a roof composed of a number of small spans. If a house, say 45 feet wide, be required, the lean-to form, even with the lowest practical pitch, would cause the roof to be inordinately high. A single span-house, with the same pitch, would reduce the height of roof one-half, but even then it would, probably, be much higher than is convenient for either growing or archi- tectural purposes. It might be, under these circumstances, doubtless advantageous to split up the roof into say 3 spans of RIDGE AND FURROW HOUSES. 55 15 feet each, or a central span of 20 feet and two subsidiarj' side spans of i2| feet each. But neither of these roofs could with justice be called, a " ridge and furrow " roof. If, however, the roof wer^ split up into 9 spans of 5 feet each, this would then be a ridge and furrow roof proper. The above roofs are shown in Fig. 33 — viz., lean-to single span, 3 span and 9 span Fig. 33. — Vertical block section, showing modes of constructing a roof of various spans. "ridge and furrow" roofs. Of course we are now treating the roof simply in relation to ridge and furrow principles ; doubt- less a house 45 feet wide could be treated better by having a roof of a design perfectly different from any of those shown. Now it is obvious that (exclusive of the gables) : — All the above roofs have the same amount of glass area — That the smaller the number of spans, the greater the area of glass in the gables — That, so far as shape alone is concerned, the amount of solar rays admitted by the roof would vary only in an in- significant degree if it were composed of one, three, or nine spans, provided the pitch and aspect remained the same — Also that, the greater the number of spans, with the greater ease can a uniform distance be maintained between the plants and the glass. But the greater the number of spans, the greater the number of valley gutters and ridges. Consequently the larger is the area of obscurity formed by them, and the greater are the constructional difficulties connected with keeping these 56 GROWING-HOUSES. valley gutters water-tight and in good repair, and in properly supporting the roof. Therefore we may safely conclude that, given the necessity to treat a roof on the span principle, a ridge and furrow roof proper admits less light and is more difficult of construction and maintenance than a roof having the smallest practicable number of spans. When Sir Joseph Paxton desgined the large conservatory at Chatsworth and also the 1851 Exhibition, now the Crystal Palace, it was considered that he had solved a great problem in connection with horticultural buildings, but subsequent ex- perience has proved that the ridge and furrow mode of con- struction, however it may be adapted for railway stations and large buildings, presents, at any rate for ordinary horticultural, work, more disadvantages than advantages, especially when cost of maintenance is of consideration. IRON V. WOOD HOUSES. There can be no doubt that a building constructed of iron requires a smaller bulk of material to possess a given strength than one constructed of wood. Therefore the framework of an iron house cannot obstruct so many solar rays as a correspond- ing one of wood. This is especially the case when the sun is shining on the roof obliquely, for the shallow iron bars and rafters will throw a much smaller shadow than the thicker and deeper wood bars and rafters. Then, again, iron houses will last longer than wood, provided the material is kept well pro- tected by paint, from the action of the atmosphere. If not kept well painted, and if allowed to rust, they will not only wear out rapidly, but the rust drip produced by the condensation of ' moisture will be very disastrous to plants. On the other hand, iron houses are more costly than corre- sponding houses of wood, as iron is not worked with so much IRON V. WOOD HOUSES. 57 facility as wood ; ventilators of iron houses are frequently found unsatisfactory, and not sufficiently tight to enable the house to be well fumigated, especially in cheap iron houses. Ordinary- sized conservatories (not, of course, winter gardens or other large structures) are not capable of such effective architectural treatment in iron as in wood, as those made of the former material are apt, unless constructed inordinately massive, to look " wiry." Glass is frequently found to break much in iron houses, owing, so some people say, to the expansion of the iron. But it will be found that this defect is generally owing to the expansion of the glass against the non-resisting metal, rather than to the expansion of the metal itself, especially if the glass has been cut, in the first instance, to fit accurately between the sash bars. The linear expansion by heat from 32'^ to 212° Fahr. is Of Glass, I part in 1161. „ Cast Iron, I ,, ,, 889. ,, Wrought „ I ,, ,, 819. So that if a pane of glass be 10'' wide, and a wrought iron sash-bar yi" thick, the pane of glass will expand -00086 1 inch and the sash-bar only "000015 i'l'^h on the temperature being raised from 32° to 212°; or in the ratio of 861 to 15. Of course there is the expansion and contraction of tie rods or constructional parts by which the house is held longitudinally. So that there may be various strains causing the panes to deviate from their rectangular shape, when of course breakage will occur. But provided iron houses are constructed in a proper manner, fitted and erected by experienced men, and the glass put in with a certain amount of "play," no difficulty will be found to exist by glass breaking. With thoroughly good work in iron houses, ventilators and other adjuncts will be certain to fit well ; but with cheap, inferior work, iron houses will prove a continual source of annoyance. Where curvilinear work is con- 58 GROWING-HOUSES. sidered necessary or desirable, iron bars and ribs may be judi- ciously used instead of wood. Iron is a good conductor of heat (the ratios being : — iron 450, glass 14, deal 3), conse- quently an iron framed hot-house would radiate internal heat more rapidly than a wood framed house. We have hitherto been speaking; of houses in which the frame- work is wholly constructed of iron. A compromise is, however, not only possible, but frequently judicious, and some of the best growing-houses to be seen are those which have been con- structed of a combination of wood and iron. These have some of the parts most difificult and costly to work, when made en- tirely of iron (such as plates, sills, mullions, posts, &c.), con- structed of wood. The rafters are also of wood, much lighter than usual, but made amply strong by light iron tie rods. The intermediate vertical and roof sash-bars, the purlins, also some of the hanging lights, are made of light but effective sections of L and T iron. In this way very strong, durable, and light houses can be constructed, admitting the maximum amount of solar rays and casting very little shadow, no matter at what ob- liquity the sun happens to be shining. Independently of any other considerations, where first cost is of importance, wood houses, or a combination of wood and iron, is to be recommended in preference to those entirely of iron, as it is not advisable to build the latter unless they are carefully and accurately fitted; and careful, accurate fitting, no matter what labour-saving contrivances may be employed, means extra expense. Wood houses, if properly constructed and kept well painted, are very serviceable, and will last a long time. WALL-TREE PROTECTORS. A wall affords a certain protection, contingent more or less upon its aspect, to trees growing against it ; but this protection WALL-TREE PROTECTORS. 59 is not always sufficient. A coping of some material placed on the top of the wall has been found not only effectual in pro- tecting the trees from some of the disastrous consequences of autumnal rains, but to a great extent from late spring and autumn frosts, and also in checking the upward current of air heated by the warm wall, which upward current would other- wise rapidly convey away by convection the heat from the wall. Even a simple coping, however, unless it be very narrow indeed, may not be advisable at all seasons, especially when the trees do not require to be kept from summer dews. On the other hand, when the fruit and wood are ripening, protection from wasps, birds, &c., as well as from cold and wet, is found very advantageous. In other words, the conditions to fulfil are : — cheapness ; extreme portability ; power of partial or entire ven- tilation ; no skilled labour or excessive time occupied in erec- tion, manipulation, or taking down. To fulfil all these conditions various schemes have been adopted. Permanent copings of Fig. 34. — Vertical section of wall-tree coping and cover. Stone or cement have given way to temporary copings of wood, &c. Tiffany, netting, and other materials have been employed to hang from the coping in front of the wall. Horticultural builders have brought out portable glass copings, supported by 6o GROWING-HOUSES. brackets bolted to or through the wall (see Fig. 34), or by posts from the ground in front, the glass on the coping, or the whole coping itself, being easily removable when desired. The neatest and most efficient coping of this description is shown by Fig. 35. Fig. 35. This coping is constructed with lights which are made to slide in the brackets and are held in position by wedge clips, so that a long length of the glass coping can be taken down or put up in a few minutes. This is a far superior plan to those which involve handling single panes of glass, for it is frequently necessary to take advantage of a shower, rather than trust to the clumsy substitute of syringing. Then there has been vertical glass and net protection, formed by upright portable glazed frames or wire netting, supported between the bottom plate of the coping and a sill resting on iron shoes or short posts iixed in the ground. The late Mr. Rivers invented an ingenious form of wall-tree cover of which the design has been registered. (Exclusive manufacturers, Crompton & Fawkes, Chelmsford.) Upright pieces of wood are placed at distances of about 24 inches apart. Horizontal and diagonal grooves having been cut, glass WALL-TREE PROTECTORS. 6l IS slid in them between the uprights, and forms the necessary protection. If a small amount of ventilation be required, the horizontal strips of glass are taken out. Pieces of perforated ^^^ VENTILATOR Fig. 36. — Interior perspective of Rivers' wall-tree cover. Fig. 37.— Vertical section of front of Rivers' wall- tree cover. zinc slipped in their place will prevent wasps, &c., entering. ' If the whole or part of the front requires to be open, as much of the glass as is necessary can easily and quickly be taken out, 62 GROWING-HOUSES. and as easily and quickly slipped back when required. This is a simple form of wall-tree cover ; but as it involves handling very many separate panes of glass, a tedious and troublesome operation, it does not possess the advantages of the wall-tree cover shown by Fig. 38. This structure (designed by the author) consists of a light permanent framework in front, slightly inclined from the per- pendicular, held in its position by light ribs, which can be fixed to the wall by bolts and nuts ; or, where there is an objection to anything passing through the wall, by coach screws. The roof ribs are hollowed in the centre to carry to the front any drip or rain which may find its way between the roof sashes. Upon the roof and front are lights held by special hinges, so constructed that the lights are secure, and yet may be unhooked in a moment when it is necessary to do so. Each light is pro- vided with a simple ratchet set open, so that ventilation to any extent may be effected from the inside or out, and the lights are held automatically in whatever position they may be placed, or the cover can be stripped off upper or lower lights, or both, in a few moments. Nets may be hung in place of upper or lower lights, or both. Or nets may be hung behind the lights without in any way interfering with their action. The lights, when not required for the cover, can be used for cucumber or melon frames, or for protecting bedding plants ; and in any case the front part of the structure may he used for growing pot strawberries, small salad, &c. This orchard house possesses, amongst others, the following advantages : — Power to check, when necessary, the upward current of air, so that a wall warmed by the sun may not too rapidly part with its heat ; protection from heavy rains, descend- ing or horizontal currents of cold air ; protection from wasps and birds ; power to give easy, rapid ventilation in any part to any extent, even to that of stripping the wall of any protection WALL-TREE PROTECTORS. 63 3 64 GROWING-HOUSES. whatever ; durability ; lightness ; extreme portability ; complete absence of complication, so that no skilled labour is required in its erection, manipulation, or removal. Permanent wall copings are usually made to project not more than about 6" over the wall. Temporary wood copings are made to project 9", 12", or even more. Glass copings are made i' 6" to 4' o" wide, and these copings with combined vertical protection up to 6' o" wide. In any case, permanent copings cannot be recommended. Protectors for wall fruit, as described above, can be con- structed in various ways, and may prove very useful. But it will generally be found that at best they are a makeshift, and that with a trifling addition to the first cost, an efficient and serviceable lean-to house can be erected, by which, even with- out, but especially with, a heating apparatus, results of a more satisfactory and reliable character can be obtained. In this case reference should be made to the section " Lean-to Houses," p. 41 et seq. COMBINATION OF BUILDINGS. The futility of attempting to treat this part of the subject in detail is conclusively shown when we consider that, according to the laws of permutations, with, say twelve conditions, no fewer than 479,001,600 different changes can be obtained There are a great many more than twelve conditions which go to make up the most advantageous combination of horticultural buildings, therefore we can only indicate here what some of those conditions are, give a few hints regarding them, and annex one or two illustrations of combinations which may be service- able under certain circumstances. The arrangement of any combination of buildings will depend upon, amongst other things : — The size of the establishment; COMBINATION OF BUILDINGS. 6S the amount of money to be devoted to the purpose ; what is required to be grown ; if fruit or flowers must predominate ; whether space has to be devoted chiefly to specimen plants or general purpose produce ; whether the luxuries or the neces- saries of fruit are required ; whether the produce is chiefly re- quired for show or utility ; whether in or out of season ; whether much or little forcing is necessary ; the tastes and habits of the proprietor ; the exigencies of the site and constructional re- quirements. Not only will these and other conditions determine the class of buildings to be erected, but more especially the proportions which the various parts of a combination must bear to one another. A wall may already exist, against which it is expedient to erect a range of buildings. No wall may exist, but it may be considered advantageous to erect one, in order, not only that a lean-to range may be built against it, but to provide an avail- able north wall for boiler-house, potting-shed, &c. Or, again' no wall may exist, and it may be necessary for the combination of buildings to take the low span type, in order that a high wall or similar obstruction may not be required. In any case, horticultural building;s should be planned so that they are compact, not straggling ; so that the buildings for consecutive operations may be arranged, as nearly as possible, in consecutive order ; so that there may be no long, undivided houses ; so that the boiler is in a convenient position for the work which it has to do ; and so that each separate building Fig. 39. — Vertical block section of combination of span-houses. (Faulty. ) does not suffer in efficiency through being placed in combina- tion with others. If a certain combination is desired, but the 66 GROWING-HOUSES. expense of building such combination at one time is too great, it is better to build part at first, and leave the remainder till later, rather than lessen the first outlay by decreasing the effi- ciency of the whole combination. A plan favoured by some is to build span-houses together, so that each of the inner walls is common to two houses. (See Fig. 39.) COMBINATION OF BUILDINGS ^J In this manner, if three houses are built, of course two vertical walls are saved ; but it must be remembered that the inner house can have no bottom ventilation along the sides, and that each of the two outer houses can have bottom ventilation only along one side. Where a combination of lean-to or three- quarter span-houses is built against a south wall, it is frequently advantageous to have span-houses meeting the combination at right angles ; but these span-houses should not be so long nor so high as to cause solar obstruction to the other houses. A useful combination for a small garden is shown by Fig. 40, viz., a range 45 feet long. This range is intended to represent a central three-quarter span-house, 15 ft. long by 12 ft. wide, by 6 ft. high at front, and two side lean-tos, eacli 15 ft. long by 10 ft. wide, and 5 ft. high at front, the brick wall at back being one uniform height of i o ft. above the ground. The building may be used as two vineries and a plant-house ; or vinery, peach-house, and stove ; or vinery, melon-house, and greenhouse ; or stove, greenhouse, and vinery ; or cucumber-house, stove, and greenhouse ; or peach-house, cucumber-house, and stove ; or fernery, plant- house, and orchid-house; or as various other "combinations; depending, of course, upon the interior fittings, heating appara- tus, &c. Fig. 41 shows a combination which may be suitable for a moderate-sized garden : — This consists of a central three-quarter span and two lean-to houses. The wall-plate at back will be at the same level through-out, if the projection of centre house in front, and perhaps its higher eaves, are allowed, as in Fig. 40, to compensate for the decreased proportional height at back of it. In addition to this range, there may be isolated rows of pits in front, part heated for cacumbers, &c., the 68 GROWING-HOUSES. remainder unheated for bedding stuff and other general purposes. N N Fig. 41 shows a ground plan of a combination suitable for a larger establisliment : — This was a combination designed by the author for W. H. Stone, Esq., of Lea Park, Godalming, and (with some modifi- cations) carried out. It will be seen that a wall has against it COMBINATION OF BUILDINGS. 69 '^ B n ts > t-l v> i-i w re •a a ' 3 3 • - 9 a Si iTl : 1 ' o O W "^ t^ S ^ O " S S- S e £L "< ^ 3 re S rt 3* E^ "• <- S p- g °- (3 s^ ." " n * ;; ?r •n B n 5 •-^ i> » B HO SI ?i (n d ^ 1— 1 s 1 [r a-E