426 THE SOrECES OF AECHITECTUEAL TYPES. 



but we may trace some kinsliip between each kind of irreg- 

 ularity and the local circumstances. We see the broken 

 rocky masses amid which castles are commonly placed, mir- 

 rored in their stern, inorganic forms. In abbeys, and such- 

 like buildings, which are commonly found in comparatively 

 sheltered districts, we find no such violent dislocations of 

 masses and outlines ; and the nakedness appropriate to tte 

 fortress is replaced by decorations reflectiug the neighbour- 

 ing woods. Between a Swiss cottage and a Swiss view 

 there is an evident relationship. The angular roof, so bold 

 and so disproportionately large when compared. to other 

 roofs, reminds one of the adjacent mountain peaks ; and 

 the broad overhanging eaves have a sweep and inclination 

 like those of the lower branches of a pine tree. Consider, 

 too, the apparent kinship between the flat roofs that prevail 

 in Eastern cities, interspersed with occasional minarets, and 

 the plains that commonly surround them, dotted here and 

 there by palm trees. You cannot contemplate a jDicture of 

 one of these places, without being struck by the predomi- 

 nance of horizontal lines, and their harmony with the wide 

 stretch of the landscape. 



That the congruity here pointed out should hold in 

 every case must not be expected. The Pyramids, for ex- 

 ample, do not seem to come under this generalization. 

 Their repeated horizontal lines do indeed conform to the 

 flatness of the neighbouring desert ; but their general con- 

 tour seems to have no adjacent analogue. Considering, 

 however, that migrating races, carrying their architectural 

 systems with them, would naturally produce buildings hav- 

 ing no relationship to their new localities ; and that it is 

 not always possible to distinguish styles which are indige- 

 nous, from those which are naturalized ; numerous anoma- 

 lies must be looked for. 



The general idea above illustrated will perhaps be some- 

 what misinterpreted. Possibly some will take the proposi- 



