782 EXAMINATIONS passed in is often only so much " book " learnt almost by heart ; with those who do really well the case is different. The value of these examinations is only that they show that men can apply their minds, and can express themselves passably well. The subjects should be chosen much less for their value as information than for their requiring the exercise of thought. Pass men are apt to reduce all they can to the action of memory; hence subjects should be taken which require something more than memory. To detect " parrot-work," the examiners should be familiar with the text-books from which the subjects are learnt, and therefore such examinations should be in connexion with set courses of teaching. Translation of unseen passages, in Latin for instance, should be insisted on, but a diction ary might be allowed. Questions in geometry should be set in such a way that they cannot be answered by writing out Euclid by heart. The difficulty of a pass examination depends both on the number of compulsory subjects it contains, and on the standard maintained in each. Feeble men can get through an examination in one or two subjects at a time, if the standard be moderate. Thus an examination which can be passed piece-meal, like the Cambridge " Little-go," is a poor criterion of brains, while an examination embracing many subjects ensures a certain strength of head, but not lasting knowledge of any one thing. When an examina tion has to be extemporized in order to ascertain whether candidates have heads on their shoulders, it will be sufficient to read over to them once or twice some short narrative or argument, or a correspondence on business matters, and to call on them to give an account of it on paper. This will test sufficiently well many of the qualities which go to make an efficient subordinate. It remains to say something as to practical methods of examining. Originally examinations were conducted viva voce, and they still are so in part. Examinations in experi mental philosophy and natural science are valueless without something of the kind. The student must perform experi ments and explain them, and must identify and describe specimens. Viva voce examination is not well adapted for discriminating between candidates who are nearly equal, because they have not the same questions put to them, and nervousness is a disturbing element. The value of viva voce lies chiefly in detecting shallow knowledge. It convicts an impostor. On paper a candidate may avoid a searching question ; in viva voce he has no escape. The objection to its employment is its great expense. It requires very skilled examiners, two of whom ought to sit together; and the examination should last a quarter of an hour for each man. When the numbers are large this involves a long period of examination and great cost. The German system of giving only one question in each subject for a pass examination, and allowing plenty of time, but requiring a very full and perfect answer, is well suited for fairly pre pared men, who have only to be roughly classified as "excellent," "good," "fair," and "indifferent." This forces the candidate to study the whole subject carefully, while if a dozen questions are given, as in England, can didates will speculate on passing with a knowledge of only half the subject. Essays may be used in examinations in two ways. Sub jects of a general nature, like a maxim or topic of the day, may be proposed, in which case readiness and fertility of ideas are tested, but a kind of superficiality and glibness is engendered; or the student may be required to write on some subject belonging to his course. The classical student, for instance, might write on a point of Greek history. A dis sertation written at leisure is an excellent means of judging of qualifications, and may be used for those who are past the proper age for examination. In marking a paper the examiner distributes his marks to the questions according to the difficulty or the time they take to answer. The aggregate of the marks may not coin cide with his impression, and it may be well to keep back one quarter of the marks, to be allotted afterwards, accord ing to the impression obtained when the papers are read over again, not question by question, but as wholes. It may be well sometimes to use negative marks, as an answer may reveal such ignorance as to show that some of the correct answers were " parrot-work." When different subjects are compared, a little knowledge should go for nothing, and excellence should count for much. It is a good plan to add to the marks got the excess above half the full value assigned to the paper, and then deduct one quarter of the full value, e.g., if the full value be 500 and the candidate obtain 400, his score will stand thus : 400 + 150-125 = 425. Candidates for honours maybe arranged in order of merit, as is common at Cambridge, or alphabetically classed, as at Oxford. In the first case brackets should be used, so as to class as equal those who fall within certain limits of uacer- tainty. These limits will be wider where there is room for difference of opinion among the examiners, as in composition or philosophy, than in mathematics. If the candidates whose marks differ by as much as twelve per cent, are bracketed together, we come to something like an alpha betical arrangement in classes. When the alphabetical system is adopted those who are sure of a first class are freed from anxiety. But many are in suspense aboul their class, and the difference between being in a first 01 second class alphabetically arranged is greater than that between being last in the first class or first in the second class, where the lists are in order of merit. Out of 1000 young men who come to a university with a view to taking a degree, we find from experience that, roughly speaking, the following proportions will hold good : 250 will have both good abilities and the requisite power of will, and will take creditable honours; about 200 more will be comparatively weak in one or other of these qualifications, but n?ay still get a place in an honour school or tripos; the next 150 will be the more vigorous pass men, who will show intelPgence in subjects of but moderate difficulty, will enter keenly into the life of the place, and will pass their examinations respectably; 200 more will pass without failure; the 100 that follow will meet with failures more or less frequently; and the remaining 100 will never pass any university examination at all. Some of these last instances may almost be regarded as cases of disease, arising from infirmity of will or the want of the power to fix the attention. Neglect of the early acqui sition of good mental habits is the cause of many failures. A youth may be rejected once from love of amusement or from underrating the examination, but he does not fail again if he can help it. A second failure shows moral or intellectual incapacity. On this subject see " Remarks on State of Education at Cam bridge," in Dr John Jebb s works, 1774 (here ve find the first plai< for examining the pass men) ; Peacock, On the Statutes of the Uwi vcrsity of Cambridge, 1840 ; "Whewell, Of a Liberal Education, 1848; Reports of her Majesty s Commissioners on Oxford, 1852, and mi Cambridge, 1854 (in the latter see the evidence of Dr Philpott, Prof. Stokes, Dr Merivale, Mr E, Leslie Ellis, and Mr W. Hop kins) ; Suggestions on Academical Organisation, Mark Pattison, B.D. (referring to Oxford) ; L. Wiese, German Letters on English Education, translated by L. Schmitz, 1877 ; Education in Oxford : its methods, its aids, and its rewards, James E. Thorold Rogers ; Conflict of Studies, I. Todhunter, F.R.S., 1873 ; Higher Xchook awl Universities in Germany, M. Arnold, 1874; On tlie Action of Ex aminations, H. Latham, 1877 ; Report to the French Government on Edit cation -in England, by M. Demogcot and M. Montucci, 1870; Third Report of Royal Commissioners on Scientific Instruction, 1873 ;
M. Burrows, Pass and Class, Oxford, 1873 ; Student s Guide U