beeu constructed, as far as could be, with all the sides and all the 

 roof of that material. 



It is my purpose to show that this maximum quantity, so far 



recollected, is eminently diffusive, so that, as illustrated in the 

 Pantheon at Rome, a single foot of glass is found to illuminate 

 in an agreeable manner 3000 cubic feet of space ; and the 

 attention devoted in modern times to the subject of picture 

 galleries, has enabled architects to remove, by a moderate em- 

 ployment of glass, shadow from every part of the internal surface 

 of large rooms and galleries, from which side light is altogether 

 excluded. 



The late Mr. Loudon, whose acquaintance with such subjects 

 is well known, states that every superficial foot of glass is capable 

 of cooling one and a quarter cubic feet as many degrees per 

 minute as the degree of the external atmosphere falls short of 

 that within a building. To make a house, therefore, as I have 

 said above, with all the sides and all the roof of glass, is virtually 

 to construct a frigidarium of which the cooling powers throughout 

 a large proportion of the year can only be overcome by an 

 immense supply of artificial heat. It must be unnecessary 

 therefore to insist on the value of any plan by which the 

 admission of light and the retention of heat may be to the largest 

 practicable extent secured. 



The optical law of reflection is, that any incidental ray of light 

 is thrown from a reflecting surface at a corresponding angle to 

 its own, but in the opposite direction. It will not therefore be 

 difiacult to conceive the possibility of so placing opaque reflectmg 

 surfaces (I do not necessarily contemplate or imply polished 

 planes or glass mirrors) in a building as to prevent the escape 

 and throw back into the interior the more vivid southern rays, 

 and thus to increase the positive amount of light within such 

 building; and in so far as those reflecting surfaces may be 

 formed on a cellular system, or with materials unfavourable to 

 the conduction and loss of heat, the economisation of that 

 valuable agent will be also effected. 



The means by which this great object may be secured are 

 fortunately of the most simple and inexpensive character. I will 

 suppose, for example, that it is necessary to erect a house for 

 the reception of plants, in the form of a long parallelogram, whose 

 chief dimension shall run north and south. It will be a question 

 in such a case dependent on the value accorded to a due east 



