LETTERS. 221 



of Aristides, you of course remembered that the former 

 had the largest scope for action, and that he influenced 

 his countrymen to fall into all his plans, while they 

 banished his competitor, not by his superior wisdom or 

 goodness, but by those intrigues and factious artifices 

 which Aristides would have disdained. Themistocles 

 certainly did use had means to a desirable end : and if 

 we may asume it as an axiom, that Providence will for- 

 ward the designs of a good sooner than those of a bad 

 man, whatever inequality of abilities there may be be- 

 tween the two characters, it will follow that — had Athens 

 remained under the guidance of Aristides. it would have 

 been better lor her. The diiference between Themis- 

 tocles and Aristides seems to me to be this : that the 

 former was a wise and a fortunate man, and that the 

 latter, though he had equal wisdom, had not equal good 

 fortune. We may admire the heroic qualities and the 

 crafty policy of the one ; but to the temperate and dis- 

 interested patriotism, the good and virtuous dispositions 

 of the other, we can alone give the meed of heartfelt 

 praise. 



I only mean by this, that we must not infer Themis- 

 tocles to have been the hetfcr or the greater man, because 

 he rendered more essential services to the state than 

 Aristides, nor even that his system was the most judi- 

 cious, — but only, that by decision of character and by 

 good fortune, his measures succeeded best. 



«•• v^ * * 



The rules of composition are, in my opinion, very few. 

 If we have a mature acquaintance with our subject, there 

 is little fear of our expressing it as we ought, provided 

 we have had some little experience in writing. The first 

 thing to be aimed at is perspicuity. That is the great 

 point which, once attained, will make all other obstacles 

 smooth to us. In order to write perspicuously, we should 

 have a perfect knowledge of the topic on which we are 

 about to treat in all its bearings and dependencies. We 

 should think well beforehand, what wiU be the clearest 

 method of conveying the drift of our design. This is 

 similar to what painters call the massing, or getting the 



