ON THE INFLUENCE OF EXAMINATIONS. 443 



general scheme of reform it would be well to devote more of the college 

 funds available for these purposes to scholarships for undergraduates who 

 have already begun residence, and especially to post-graduate fellowships 

 and studentships for research. An increase of the endowments of 

 secondary schools to enable them to award more leaving scholarships 

 tenable at a university would be an etlicient substitute for the present 

 system of open entrance scholarships at the colleges of Oxford and Cam- 

 bridge. Any such change would, however, require the co-ordination of 

 the whole secondary education of the country. 



In arranging an open entrance scholarship examination in such a 

 subject as natural science the chief difficulty is to provide for two 

 distinct classes of candidates: (1) boys, often under eighteen years of 

 age, just leaving school ; and (2) those, usually rather older, who have 

 spent some time at a university or technical college, specialising on the 

 work in which they are to be examined. In order to help the school- 

 master to give a thorough training in the groundwork and main principles 

 of science, it is advisable that the papers set in the scholarship exami- 

 nations should largely deal with these parts of the subjects ; while, to 

 properly test the merits of older candidates wlio have spent some time 

 at the work, more advanced questions are requisite. A satisfactory 

 judgment cannot be formed on the results of the examination alone, 

 and under present circumstances it is necessary to make allowance 

 for what is known or can be ascertained of the antecedents of the 

 candidates. 



U 10. University entrance scholarships, while successful in so far as 

 that they do pick out the able students in each subject, are at present 

 doing great harm by encouraging early and excessive specialisation to the 

 detriment of the student's subsequent career. Thus, for their knowledge 

 of chemistry and physics, scholarships are awarded to Iwys of eighteen 

 who have in far too many cases a very inadequate grounding in mathe- 

 matics, are ignorant of history and of modern languages, possess a 

 smattering of Latin, and cram up subsequently enough Greek to carry 

 them through the ' Little Go ' or its equivalent. Equally bad is the case 

 of the winner of a classical scholarship who, beyond his knowledge of 

 Greek and Latin, has a slender aco[uaintance with Euclid, algebra, ai-ith- 

 metic, and French. 



Worse still is the condition of the mathematician who as regards 

 general education is more poorly equipped than the rest. 



The best of these men often repair their deficiencies later by their 

 own efibrts ; the second best remain losers. 



This evil might be met by insisting that all scholarship candidates 

 should pass a suitable matriculation examination before they were 

 allowed to compete for scholarships in special subjects. Most matricu- 

 lation examinations would, however, require to be considerably improved 

 and widened in their scope before they could be used for this purpose. 



U 11. Whilst the Oxford and Cambridge examination of schools seems 

 to me to have done unmixed good, I hold that open examinations for 

 college scholarships have done, are doing, and will continue to do much 

 harm by encouraging schoolboys to specialise early in some one branch, 

 whether of literature or of science. The schoolmaster is compelled («) by 

 the natural desire to advertise his school, (h) by the absolute necessity of 

 meeting the reasonable wishes of parents, to prepare liis boys for open 

 college scholarships, obtainable only by candidates under nineteen years 



