178 ON TASTE IN ORNAMENTAL DESIGN. Paet II. 



plain how in its defective parts such improvements might be 

 substituted as would make it perfect. There are always many- 

 bad designs, and the reason should be stated why they are so ; 

 there are others that have perhaps one good feature, and it 

 should be pointed out; there are others which are almost 

 good, but yet, through some imperfection, are prevented from 

 being so ; and it should be explained how such and such a 

 change would make the required improvement. It is this 

 very power of perceiving the beautiful, however small in 

 quantity, which (as I shall have occasion to show) enabled 

 the Greeks to select from, and improve on, the works of 

 others less talented than themselves, and we need not be 

 above the attempt to follow so good an example. May we 

 only succeed in it ! Such lessons would convey more prac- 

 tical instruction than the mere examination of the most per- 

 fect design. It is easier to teach by showing why something 

 is not good, and how it might be made so : and this negative 

 process will afford to beginners the best of lessons, care being 

 taken at the same time that the most perfect and beautiful 

 objects be constantly recommended to their attention. For 

 while instruction may be conveyed by teaching them what to 

 avoid, the general taste should be guided by the contempla- 

 tion of what is good ; one appealing to the reasoning faculty, 

 the other educating through the eye. 



6. [It is of the highest importance for creating and im- 

 proving the taste of the public that good designs should be 

 constantly before them. Even those who are gifted naturally 

 with a certain amount of it are liable to have it vitiated by 

 frequently seeing bad models ; and a more cultivated taste is 

 occasionally warped and led insensibly towards the extrava- 

 gant if, debarred from the contemplation of the beautiful, it 

 has faulty designs constantly before it. Many, again, allow 

 themselves to be deceived by the fiat of men whose names are 

 thought to give weight to their opinion ; and it is thus that 



