SECTION V. ■:.ai7qi.3qoi-v 



:•;, ,T":'- ■■■■ I 



MATERIALS OF C N S T R U C T I p N,..,i ., , , ... ,,;,.-^ 



1. W(yfk7nanship. — However excellent and adaptable may- 

 be the design of a horticultural erection, if the work be badly 

 executed the structure will generally be defective in the work- 

 ing, and the trouble of management will be greatly increased.' 

 Bad foundations, bad roofs, bad-fitting sashes, rendering them 

 difficult to open and shut, bad glazing, and bad workmanship of 

 every description, are too common to exist without being a very 

 perceptible evil, and one that is much complained of by practical 

 gardeners, upon whom the consequences of this method of coil-' 

 struction generally fall. In all regular work, coming under the' 

 province of the architect or engineer, there is generally particu--^ 

 lar attention directed to the facility of working, and ingenuity' 

 is exerted to its utmost limits to perfect and simplify those 

 facilities, however temporarily the structure or work may be 

 constructed. But horticultural buildings, relatively to civil 

 architecture, appear to be dn anomalous class of structures, not 

 coming strictly within the province of 'the architect, — except 

 in so far as they may be related to the house in an architectural 

 point of view, — and hence they are more the subject of chance 

 or caprice in design, ' an'd of local convenience in execution, 

 than any other department' of rural architecture. The subject 

 of horticultural architecture has not been deemed of sufficient 

 importance to induce civil architects to make themselves "aci^ 

 quainted with the' ' principles on which plant-houses should be* 

 constructed, or to consider the nature of workmanship in relation 

 to its work ; arid, consequently, the construction of horticultural 

 buildings is either.,left,; wholly, to- gardeners, who understand 

 little of the science of architecture, or whoUy^to architects, who-, 

 understand as little of the science of horticulturei.^^j.;Tiae consewi 

 quence, in either case, is generally incongruity itt kppearancO'/J 

 9^ 



