November 21, 1901] 



NA TURE 



69 



of reference of the temperatures is unusual. The temperature 

 for any day is not referred to the ascertained 25-year mean for 

 that day, but to the corresponding point of the first component in 

 the harmonic resolution of the 25-year curve, and it is the 

 differences from this standard which are considered. To obtain 

 trustworthy means for the several winds the differences for each 

 ten days of the year or for each month are grouped, and the 

 winds are also grouped according to their average annual effect 

 upon temperature, which is tabulated in the paper. Thus 

 N., N.E., and E. winds (E., S.E. and S. gradients) are grouped 

 as "cold winds," S., S.VV. and W. winds as " warm," and 

 N.W. and S.E. as " temperate." 



A series of diagrams exhibits the results obtained, and 

 it is noteworthy that the chief characteristics of the 

 peculiarity which it is sought to explain, viz. maxima of 

 warming effect in January and July with maxima of coolitig 

 effect in May and November, are traceable in different ways in 

 the temperature curves for separate winds and still more 

 markedly in those for the groups of winds. Thus the peculiarity 

 is only partly attributable to the prevalence of warm or cold 

 winds ; part is due to a similar peculiarity in the seasonal varia- 

 tion of temperature of the individual winds themselves. Thus 

 the May minimum is shown to be due partly to the special 

 prevalence of " cold winds" and partly to the relative coldness 

 of those winds at that season : the corresponding November 

 minimum is attributed to the prevalence of " temperate 

 winds " and the exceptional coldness of those winds at that 

 period of the year. The July maximum corresponds to the 

 exceptional warmth of the usually cold or temperate groups of 

 winds and the January maximum corresponds especially to the 

 frequency of occurrence of " warm winds." 



The half-yearly component of the variation of temperature in 

 individual winds remains unexplained, but the following facts 

 are noted in connection with it : first, a similar effect is found in 

 the temperature variation of sea-water at stations surrounding 

 these islands and, secondly, a similar second order component 

 with similar epoch is found in the seasonal variation of the baro- 

 metric gradient between London and Valencia and still more 

 conspicuously in the barometric gradient between London and 

 Aberdeen. 



The d.ita as to the relation between wind, or gradient, and 

 temperature have been obtained for Kew only. 



The paper also contains an account of the variation of tem- 

 perature with the type of weather prevailing, whether cyclonic 

 or anticyclonic, and it is shown that the effect in question 

 cannot be ascribed to the differences of frequency of these types 

 at different seasons. 



To the paper is appended a note on the effect of sea tempera- 

 ture upon the seasonal variation of air temperature. In this 

 an attempt is made, by the application of the principle of the 

 vector composition of sine curves of the same period but with 

 different epochs, to resolve a resultant annual temperature 

 oscillation into components corresponding to the "original 

 oscillation " and the superposed effect of sea or land. By the 

 application of the principle in the case of Kew, it is shown that 

 the amplitude of the "original oscillation" at Kew cannot be 

 less than 5-3° F. and the effect of the surrounding sea corre- 

 sponds toan oscillation which cannot exceed 8'3° F. in amplitude. 

 The resultant oscillation at Kew ought, however, to be resolved 

 into three components — the original component, that due to the 

 surrounding land and that due to the sea ; but there are not 

 sufficient data to determine them directly. 



Application is also made of the same principles to the resolu- 

 tion of the temperature variation at Scilly and at a station in 

 Siberia. The numerical results are not to be regarded as final 

 on account of the inadequacy of the data used. 



T^ 



RESEARCH IN UNIVERSITY EDUCATION. 



■"HE development of higher education in the direction of 

 research was the keynote of the address delivered by Prof. 

 J. ti. Macgregor, F.R.S., at the University of Edinburgh on 

 October 15, in opening the Natural Philosophy Class as the late 

 Prof. Tait's successor. Research methods should be used in 

 education from the Kindergarten to the University ; because the 

 spirit of self-help, of inquiry and of inventiveness which they 

 encourage is at the foundation of all progress in science and 

 industry. When science began to be studied in our schools and 

 colleges about forty years ago, the schoolmen of the day followed, 



NO. 1673, VOL. 65] 



with few exceptions, the methods which they used in teaching 

 the humanities. Lectures and books provided the material and 

 examinations the test of retentivity. The system was funda- 

 mentally wrong when applied to science though sound for studies 

 ot literature. Investigation is necessary in both cases if progress 

 is to be made, but, as Prof. Macgregor remarks, " while in science 

 the outfit of the laboratory consists of apparatus and tools, in 

 language it consists of the text and the lexicon." 



The neglect of the spirit of research in the study of science is 

 largely responsible for the want of public sympathy with work 

 of investigation and the inadequacy of provision made for it. 

 Nations like those of America and Germany which have recog- 

 nised that research is not only an educational discipline but 

 exercises a powerful influence upon industrial development, now 

 take the initiative where we were once the leaders. Formerly, 

 it was necessary for the young American to go to Germany to 

 obtain the pioneer spirit, but the need no longer exists, for the 

 leading universities of the United States have been remodelled 

 on modern lines. In Great Britain the conservative spirit 

 prevails and has prevented the course of university development 

 demanded by the requirements of the age. The characteristic 

 attitudes of the German, American and British peoples are as 

 old as the prophets, from whom Prof. Macgregor derives an 

 appropriate illustration. 



" The German," he remarks, " may be said to have soughir 

 wisdom for her own sake as being more precious than rubies, . 

 and he is finding now that she has length of days in her right 

 hand, and in her left riches and honour. The American, though 

 he sought her not, heard her crying at the gates : I, Wisdom, 

 dwell with prudence and find out knowledge of witty inventions ; 

 and having heeded her cry, he too is reaping his reward. We 

 Britons have neither sought her for her own sake nor heeded her 

 cry, but have said to ourselves : There is no new thing under the 

 sun. He that increaseth knowledge increaseth sorrow, and much, 

 study is a weariness to the flesh. The sleep of the labouring man 

 is sweet. Vet a little more sleep, a little more slumber, a little 

 more folding of the hands in sleep. And now we fear that 

 poverty is coming as one that travelleth, and want as an armedl 

 man." 



It is the duty of those of us who are awake to national neces- 

 sities to exert ourselves in the endeavour to arouse the British 

 people to action, and our political leaders to a sense of respon- 

 sibility for future welfare. More liberal provision must be made 

 for the increase of knowledge, and men who devote themselves 

 to research must be prized as highly as those who have con- 

 tributed in other ways to the progress of the nation. There 

 must be increase of funds and increase of freedom in the 

 universities and the guiding principle of the work must be 

 research. Prof. Macgregor emphasises these points in the 

 following concluding part of his address. 



Research is costly. It means increased teaching staff and 

 adequate provision of . all the requisite appliances. Much good 

 work, it is true, may be done with a comparatively small outfit ; 

 but to obtain the best result, the outfit, if not lavish, must at 

 any rate be generous. And as the importance of research by 

 students has never been recognised amongst us, the present 

 outfit is meagre. 



It will perhaps occur to most of you that the princely gift of 

 2,ooo,coo/. which Mr. Carnegie has made to the Scottish- 

 people for reducing the cost of the higher education and in- 

 creasing its efficiency may be drawn upon for the present 

 purpose, and may be sufficient. Doubtless it will be drawn 

 upon, but it will certainly not be sufficient. When we think of 

 the number of colleges which are to be assisted, and of the 

 number of different departments in each, we see at once that 

 the amount which any one department may expect to receive 

 must be comparatively small. The moiety of the Carnegie 

 fund which is devoted to equipment would build, equip and 

 maintain perhaps about twenty laboratories of the more ex- 

 pensive kind such as are required in the various sciences and in 

 their numerous applications. It becomes obvious, therefore, 

 both how munificent a gift the Scottish people have received, 

 and, since each university ought to possess many of these 

 laboratories, how inadequate it is for the introduction of research 

 study into the various departments of university work. 



In Germany the nation itself provides for research, and does 

 so generously, because it is, and has long been, an investigating 

 nation. We are not ; and if we introduce research into our 

 universities it will be because, like the Americans, we have 



