20 THE LIFE OF SCIENCE 



torian must try not only to describe these evanescent discoveries 

 but to find in science that which is timeless. When he does that 

 he comes very close to the historian of art. To put it in other 

 words, a man's name may be immortalized by his discoveries. 

 Perhaps there was nothing else in him deserving of remembrance? 

 He may have been a poor sort of man, a man whose mind was as 

 sharp and narrow as a knife-edge? Or else the historian betrayed 

 him? In so far as a scientist is also an artist, his personality can 

 survive, otherwise not. It is the historian's main duty to revive the 

 personalities, rather than to enumerate their scientific excres- 

 cences. Discoveries may be important, but personalities are in- 

 finitely more so. 



V 



The materials investigated by historians of art often are of great 

 value to historians of medicine, because artistic traditions are 

 likely to be more tangible than purely scientific ones. This is espe- 

 cially true of ancient and mediaeval times, during which the dif- 

 fusion of knowledge was necessarily difficult and erratic. Beautiful 

 monuments had on the whole a better chance of survival than 

 others, and their language is easier to understand, even to-day. 

 Dr. Sigerist has given remarkable examples of the mutual aid of 

 the history of medicine and the history of art in his lecture, "The 

 historical aspect of art and medicine/' Remember his pictorial his- 

 tory of the plague, and his account of the transformation of Apollo 

 into St. Sebastian, both being saviors or intercessors in times of 

 pestilence. 



Such examples might easily be multiplied and a balanced ex- 

 planation of them would enrich, as well as fortify, our traditions. 

 I have adumbrated some of them in the first volume of 7sis — 

 apropos of the history of cultivated plants — and in my Introduc- 

 tion to the History of Science, e.g., indicating the importance of 

 the pilgrimage roads, such as the Way of St. James (to Santiago 



