108 SCIENTIST 



will, of course, be helped by his teachers and guidance 

 counselors, by his parents, and finally by friends who have 

 recently attended the institutions he has in mind. 



All that can be done here is to discuss a few general 

 principles. In the first place the American student will 

 have to choose between going to an independent liberal 

 arts college or to a college associated with a university. 

 No general rule can be laid down that will hold for every 

 case. Matters of personal taste or family tradition may 

 properly be allowed to play a considerable role. In a 

 theoretical statistical sense, the probability of getting a 

 good science education is higher in a university than in 

 an independent liberal arts college. This is true simply 

 because all universities must have at least some research 

 going on while many liberal arts colleges have little or 

 none. On the other hand, there are at least twenty and 

 perhaps as many as fifty liberal arts colleges which offer 

 a better introductory experience in science than do any 

 but a handful of universities. The Knapp and Greenbaum 

 study referred to in the last chapter shows, in fact, that 

 the best independent liberal arts colleges send a far larger 

 proportion of their graduates on to advanced degrees in 

 science than do the colleges attached to universities. Some 

 of the differences between the two sorts of institutions 

 may be worth bearing in mind when making a choice. 



The best Hberal arts colleges may not have as elaborate 

 laboratories as the best universities and they usually lack 

 the relatively small number of absolutely outstanding scien- 

 tists that give the best universities their reputations. How- 

 ever, the best college laboratories are fully adequate for 

 undergraduate work and provide research facilities good 

 enough to attract perfectly capable if not absolutely out- 

 standing teachers. The introductory science courses in the 

 independent colleges have fewer students and almost all 



