April 27, 18S8,] 



SCIENTIFIC NE\VS. 



401 



has gained experience, it is not without pain that he looks 

 back to his own early performances, only to find pedan- 

 try in one question, obscurity in another, and want of 

 balance throughout. The rules which an examiner, bent 

 upon self-improvement, will arrive at are such as these : 

 (l) That it is his duty to ascertain the knowledge, and 

 not the ignorance, of the candidates. Truly it is a mean 

 thing for the grown man, strong in the facts and theories 

 which he has expounded for years, to set pitfalls for 

 students in order that they may stumble and fall. (2) 

 That an important question should never turn upon the 

 meaning of a new technical term. The teacher may have 

 used another word, and the candidate will then have 

 nothing to say about a topic with which he could have 

 dealt effectively had the question been differently phrased. 

 (3) That an examiner must take pride in his equity and 

 moderation, never in his learning or ingenuity. 



It is not superfluous to advise an examiner to keep 

 copies of his old papers, and look at them now and then. 

 Otherwise he may repeat himself too often. One 

 examiner might be named who never sets a paper on 

 elementary chemistry without bringing in the flame of 

 a candle. Very likely it pleases him to observe how 

 successfully that interesting theme is handled by a very 

 large proportion of the candidates. 



Let us suppose the paper set and answered. Then 

 comes the looking over of the answers, a long and hate- 

 ful job. The just examiner will now find how laborious 

 and depressing is his task. He must put himself upon 

 his honour, and resolve to take no thought for his own 

 comfort or ease. The work can only be done well by 

 breaking it up into a number of separate bits, each in 

 itself perfectly simple. First assign marks to each 

 question. Don't attempt, as so many examiners do, to 

 coax the marks, until the total amounts to exactly loo. 

 It is far better to assign the marks without more than a 

 general reference to the total, and afterwards rectify the 

 figures by the method explained below. Then consider 

 each question by itself. Set down on paper the points 

 which ought to be included in the answers. Assign 

 marks to each. Break up the chief heads into minor 

 points, if you can. When you have done, reconsider 

 the values of the questions. Take a preliminary glance 

 at the questions, to see whether your scale will work 

 out well. Then reduce your figures to the exact propor- 

 tion which they ought to bear to a total of 100. The 

 Harrow mark-reducer or a common slide-rule does this 

 in a minute or two. Work through the whole pile of 

 papers, question by question, and afterwards look 

 rapidly over each set, adding marks for intelligence, 

 accuracy, good method, etc., or deducting for the oppo- 

 site qualities. Do this before you have cast up the totals. 

 Never alter the totals on account of anything you may 

 independently know of the candidates, such as that on a 

 previous occasion you had placed them differently. The 

 total should be carefully arrived at, and treated with a 

 good deal of respect. Underlining mistakes and defects 

 all the way through is a necessary precaution against 

 the bitterness which is so apt to creep over the depressed 

 examiner. It is not a bad plan to use a red and blue 

 pencil. One colour may be used for mistakes of fact, 

 the other for faults in arrangement or style. 



A conscientious examiner will now and then examine 

 himself, arranging in order of merit a set of papers with- 

 out putting a single mark on the papers themselves, lay- 

 ing them aside for a month, and then doing it all over 

 again, in order to get an independent result. The corre- 



spondence expected varies with the subject. For 

 arithmetical and many kinds of mathematical papers, 

 variations of 5 per cent, in the totals should cause sur- 

 prise. Where the papers are descriptive, 10 per cent, 

 of variation is not unusual. Of course no practical 

 effect should be given to differences within the probable 

 error of the marker. Many examiners aflect an im- 

 possible accuracy, and assign places on the ground of 

 differences in marking amounting only to i or 2 per 

 cent. 



Now for the candidate and his share of the work. 

 The common faults, besides those due to mere igno- 

 rance, are complete misunderstanding of the question, 

 careless errors in matters of fact, and defective method. 

 The remedies are great attention to the questions before 

 beginning to write, and the drafting of skeleton answers. 

 The candidate may be recommended never to begin until 

 he has carefully studied the whole paper, and never to 

 attempt a complicated question without writing down the 

 heads of his answer. It is of the highest value to practise 

 this beforehand. Precis of a lecture instead of verbatim 

 notes, abstracts of authors read, skeletons of original 

 essays are all first rate practice, not only for examina- 

 tions, but for much of the serious occupation of after life. 

 Every answer should be read quietly over before leav- 

 ing it. No time is wasted which tends to better arrange- 

 ment or more complete statement. As a rule, it is best 

 not to attempt all the questions, and to do first the hardest 

 of those which can be answered completely and well. 

 Candidates are often curious to know " whether it pays 

 to make shots," and the general belief is in favour of the 

 practice, but it is impossible to calculate chances. Some 

 examiners are indulgent towards foolish and blundering 

 answers, and only take note of positive knowledge. 

 Others are stricter, and deduct heavily where they see 

 signs of mere guessing. A really good candidate never 

 even discusses these doubtful ways. 



There are of course the widest differences, not only in 

 examiners, but in examinations. In a pass examination 

 the main object is usually to find out what a candidate 

 knows. In a scholarship examination the main object 

 is often to find out what sort of fellow he is, and many of 

 the tests are modified accordingly. A considerable range 

 of imperfect, ill-digested knowledge may get a man 

 through a pass examination. It counts almost for noth- 

 ing in a scholarship examination where there is a good 

 field of candidates. 



The outcome of the whole discussion is that examina- 

 tion is pre-eminently a field in which method and common- 

 sense tell. Speed is important, but probably overrated 

 by the candidate, except in the few cases where it comes 

 of a full, accurate, and ready mind. Cleverness in 

 "dodges "does no good, except where the examiner in 

 phenomenally stupid. Evasion of the point of a ques- 

 tion brings no luck. At best, it leads the examiner to 

 attribute to stupidity what really belongs to ignorance. 

 It is well for the candidate if the plainness of his hand- 

 writing and the clearness of his arrangement help to 

 keep the examiner in good temper. 



Ferment.ation under the Action of the Elliptical 

 Ferment. — M. G. Jacquemin [Coinptes Rendus) has 

 proved experimentally that Saccharomyccs ellipsoideiis is 

 a true, stable ferment, quite distinct from beer yeast, and 

 that it converts saccharine solutions, decoctions of malt, 

 etc., into true wines, not to be confounded with beer. 



