April 6, 191 6] 



NATURE 



121 



classical attainments? Far be it from scientific men 

 to belittle these or any other accomplishments — philo- 

 sophical, literary-, or artistic. Our contention is that 

 — along with the more advanced study of the natural 

 sciences — these other branches of learning" should be 

 treated as subjects of special education : that they 

 ought not to dominate the general education of the 

 country. So far from having any wish to kill all 

 other learning, we desire to promote all learning, but 

 that desire does not prevent us from thinking- that 

 training in science will have to take the place in 

 schools which is now occupied by Greek and Latin. 



I am aware that our opfKjnents ma}' retort we have 

 no right to assume that persons who have had train- 

 ing in science as an integral part of their education 

 are more competent to manage the affairs of the nation 

 or to carry on business or industrial operations than 

 those who do not possess this advantage. We possess, 

 however, an example of the influence of scientific train- 

 ing on efficiency in one of the largest of our public 

 departments— the Nav\-. This is admitted, even, I 

 believe, by " Maga," to be the best organised of those 

 departments ; it is certainly the one in which the public 

 places the most confidence. But the men upon whom 

 this efficiency de- 

 pends are distin- 

 guished from those 

 of all other public 

 Services in the fact 

 that their educa- 

 tion is, from the 

 beginning, purely 

 scientific. They 

 have had no op- 



portunitv for the acquisition of that know- 

 ledge of' the classics which " Maga " appears 

 to consider necessary for the making of men ; 

 yet even the boys of the Navy have again and 

 again demonstrated by their actions that the 

 scientific training which they have received 

 has not prevented them from showing of 

 what stuff they are made. 



Nor has "Maga" the right to assume that it is 

 only the classical members of our ancient universities 

 who have conve forward so splendidly in this crisis 

 of our national life. For, side by side with those of 

 their classical fellows, " stand imperishable upon the 

 roll of honour " the names of hundreds of science 

 students who have — whether " from their despised 

 studies " or not I cannot say — also " learned and 

 taught the habit of command," and many of whom 

 have, alas ! also made the supreme sacrifice. But to 

 anyone with a " sense of proportion," it must be 

 obvious that this can have nothing to do with the ques- 

 tion at issue. For in showing their readiness to give 

 their lives for their country the members of the univer- 

 sities are doing no more than is being done by millions 

 of their fellow-subjects at home and abroad. 



E. k. SCHAFER. 



of the Institution of Electrical Engineers relate to 

 numerals punched on metal. 



There are four classes of numeral characters : (i) 

 For writing, including ordinan,' script, formal writing, 

 inscriptions moulded in metal or cut in stone; (2) for 

 typography ; (3) for scales ; and (4) for punches. Script 

 demands legibility first, but gives considerable scope 

 for caligraphy. In formal writing these requirements 

 are reversed, and for inscriptions they carry equal 

 weight. Typography makes certain intricate and 

 subtle exactions for the purpose of producing balance 

 and apparent uniformity of the characters. Such re- 

 finements do not seem to be required in the case of 

 scales and punches. Scales often have to be read in 

 a bad light, and, as I have said in my paper, elegance 

 of shape is not to be disregarded altogether, but wher- 

 ever necessarj- it must be sacrificed to legibility and 

 to special restrictions in uniformity of size and thick- 

 ness of line. 



Punches for stamping numerals on metal make two 

 additional demands on the designer. The first is that 

 a character when inverted shall not be mistaken for 

 another; the second is that if the impression is imper- 

 fect it shall do its best to be legible. The first case 



small serif, and there 



consists essentially of 

 the relation of the 6 to 

 the 9, and it is so diffi- 

 cult to make a differ- 

 ence, that other 

 numerals must lend 

 their help. The i there- 

 fore should have a 

 should be a marked 



Numerals for Scales and Punches. 



Seeixg that excellently designed numerals are 

 common on the scales of instruments, and bad 

 styles are rare, I have been surprised at the interest 

 which has been taken by engineers and others in 

 the proposed numerals which were illustrated in 

 Nature of Februan.' 24. I have adopted some of the 

 suggestions which I have received in the revised set 

 here illustrated. My intention eight years ago was 

 to produce designs suitable for the sca'les of measur- 

 ing instruments and for the dials of engine counters 

 and electric supply meters, where the numerals appear 

 through holes. Most of the suggestions which I 

 have received since my paper appeared in the Journal 

 NO. 2423, VOL. 97] 



difference between the upper and the lower part of the 

 8. I propose to retain my original 6, and I offer a 9 

 with a tail curling up a little to the left, but not 

 enough to cause confusion with 8 or o. In these 

 numerals the thickness of the lir\e is 75 per cent, of the 

 height. 



The 2 in the present set is a compromise between 

 the swan-breasted one of the first set, and the acute 

 angled type. I have tried to improve the 3 by adding 

 one unit to the length of the top bar, but this has the 

 drawback that if the lower 6/ioths is lost there may 

 be a confusion with 7. The tail of the 3 is turned up 

 higher than that of the 5 or 9. The 4, 6, 7, and o 

 remain as before. For the sake of appearance the 

 down stroke of the 5 slopes i in 10, instead of i in 20. 

 Good impressions of the upper 4/ioths of these 

 numerals should be legible. If the lower halves only 

 appear, the difference between 3 and 5 is but slight, 

 and there is no difference between the lower half of 

 the 6 and of the 8. But if the lower 6/ioths are 

 visible, I think that each numeral is legible. If 

 4/7ths of the breadth are lost on either side the 

 remaining 3/7ths are legible. 



I have shown these numerals to a friend, who said, 

 "They are quite clear but quite beastly," and he 

 pointed out that the limitations do not altogether pre- 

 clude beauty of form. He directed my attention to 

 certain good models, and I have based on these a 

 2, 3, /;, and 8. The 8 has the advantage that the 

 lower half differs from that of 6. 



I shall be glad to receive any further suggestions. 



