212

width of the alae is made a fourth of it. The length ofthose which are from sixty to eighty feet, is divided intonine parts, and the width of the alae is made equal totwo of them. The width of the alae on the right and leftof atria which are from eighty to an hundred feet long, ismade equal to a fifth part of this extent. The trabes,extending from column to column of the alae, ought to bemade as high as they are wide \

When the atrium is twenty feet wide, the tablinum ismade two thirds of it: when it is from thirty to forty feet,the tablinum is half the width. The tablinum is madeequal to two parts of fi ve into which the width of the atriumis divided, when this is from forty to sixty feet. Atria ofsmall dimensions ought not to have the same proportions asthose which are very spacious; because if we were to adoptthe proportions which obtain in large atria, in those whichare of small dimensions, the alae and tablina would be tooinconsiderable to answer the purposes for which they aredesigned: and, on the contrary, if we apply the proportions ofsmall atria in building those of ample size, the alae and tablina

1 Barbaro, Galiani, Palladio and Perrault agree in the supposition that thispassage alludes to the height of the alae, or porticoes on each side the atrium; butthey are led to this by the belief that the atrium was within the house; in whichopinion the editor of the Berlin edition concurs. In our explanation of the plan ofthe houses of the Romans we have agreed with Pitiscus, that the atrium was thegreat court before the house, having on the right and left the porticoes called alae;in which were the cellae familiaricae, or rooms appropriated to domestic purposes.Vitruvius has mentioned the height of the atrium below the beams of the columnsin a preceding passage: in this hC alludes to the height of the beams themselves.The printed copies read “ Trabes earum liminares ita alte ponantur ut altitudineslatitidinibus sint aequales,” Alte may probably have been written for altae.