March 19, 1903] 



NA TURE 



46. 



imperfectly the place of a positive result. Let us see if a 

 positive interpretation is compatible with the facts. 



In fr~e of so emphatic a declaration to the contrary, my 

 opinion may seem over bold ; yet I feel no hesitation in 

 believing that the inheritance of eye-colour in these mice, 

 so far as the record reaches, was strictly Mendelian. The 

 first cross proves that when in this case an albino (pink- 

 eye) gamete, G, meets a colour-bearing (pink-eye) gamete, 

 G', in fertilisation we must expect the resulting hetero- 

 zygote, GG', to be coloured in coat, with a dark eye. When 

 these heterozygotes breed inter se, they will form on an 

 average equal numbers of homozygotes, G(j and G'G', and 

 of heterozygotes GG' and G'G. Of these, the homozygotes 

 will all have pink eyes, but while GG will have a white 

 coat, G'G' will have some colour in the coat. The hetero- 

 zygotes, GG' and G'G, will have dark eyes and some colour 

 in their coats. 



Treating GG' and G'G as identical, we thus expert the 

 ratio 



1 GG: 1 G'G' : 2 GG'. 



Therefore the most probable distribution of the 66 mice is 

 as follows : — 



i6'5 pink-eyed albinos : 165 pink-eyed with colour in coat : 

 33 dark-eyed with colour in coat, 



and the experiments gave 



13 pink-eyed albinos : 17 pink-eyed with colour in coat : 

 36 dark-eyed, (? all) with colour in coat. 



Similarly, on crossing the hybrids with albinos, we expect 

 equal numbers of GG and GG'. Therefore the most prob- 

 able distribution of the 205 mice so produced is 



I02'5 pink-eyed albinos : i02'5 dark-eyed with colour in coat, 



and the experiment gave 



1 1 1 pink-eyed albinos : 94 dark-eyed with colour in coat. 



Experiment agrees well with expectation. In what respect 

 are they discordant? 



The case is closely comparable with that found by Miss 

 Saunders in Matthiola (Rep. to Evol. Cttee. of Roy. Soc, 

 1902), when a white hoary form crossed with a white 

 glabrous form gave purple hoary offspring ; and with the 

 production by Kdlreuter (confirmed by Gartner, Naudin 

 and Godron) of purple flowered hybrids by the union of two 

 white flowered Daturas, D. ferox and D. laevis. Why in 

 these cases the heterozygotes are atavistic we do not know, 

 but the problem need not be insoluble. 



Anyone conversant with Mendelian phenomena can now 

 predict the eye-colour of the future offspring of the various 

 unions with approximate accuracy. Pending further experi- 

 ments, we cannot predict the particular colours which will 

 appear in the coats, and for various reasons we should 

 perhaps be cautious in declaring that ail the dark-eyed mice 

 must show colour in their coats. 



From incidental comparisons of these new facts with the 

 simpler results of von Guaita an inexperienced reader might 

 suppose that the two sets of experiments had been com- 

 parable and had given discrepant results. It would have 

 prevented misconception if the author had stated that while 

 the waltzing mice he used were pale fawn and white with 

 pink eyes, von Guaita's were black and white with dark 

 eyes. It is a feature of the Mendelian view of heredity that 

 different specific results may be expected when different 

 specific materials are introduced. W. Bateson. 



Grantchester, Cambridge, March 11. 



University Education in the United Kingdom and 

 Germany. 



With reference to the admirable article on " The Uni- 

 versity in the Modern State " which you published in your 

 issue of March 12, may I point out that the figures you give 

 deal only with university education in Germany, in arts, 

 pure science, law, medicine, &c. ; they do not include the 

 very large expenditure on technical education of university 

 rank. Both in that country and in the United States such 

 education is given in technical institutions existing side by 



side with universities, but free from ordinary academic 

 control. 



Were these figures added, the comparison you make be- 

 tween the sums devoted to higher education in this country 

 and in Germany would be still more striking. For — to take 

 a single case — side by side with the well-endowed University 

 of Berlin, there is the Technical High School at Charlotten- 

 burg, which is one of the finest technical institutions in the 

 whole world, liberally supported by the State, and, notwith- 

 standing the jealous protests of the Prussian universities, 

 wisely authorised by the German Emperor to grant degrees 

 in the main branches of technical knowledge. 



J. Wertheimer. 



Merchant Venturers' Technical College, Bristol, March 16. 

 [The series of articles referred to of set purpose deals with 

 universities only. — Ed.] 



Hygrometric Determinations. 



During the past week I have made some determinations 

 with regard to the humidity of the atmosphere at this place, 

 and the results, which show a very rapid change, are, if 

 not perhaps unusual for our climate, certainly somewhat 

 striking. 



The tests were made with dew point instruments, and I 

 append the figures deduced from my determinations on four 

 days. 



(1) February 12, 

 11.30 p.m. 



(2) February 13. 



II p.m. 



(3) February 15. 

 12.30 p.m. 



(4) February 16 



1. 30 a.m. 



Case. 



NO. 1742, VOL. 67] 



The figures in the last two columns are deduced from the 

 Greenwich observations, while for those in columns (c) and 

 (d) I am indebted to Regnault. 



Unfortunately, I have not the height of the barometer at 

 the time of my determinations, but have assumed a mean 

 pressure, covering the four days on which they were made, 

 equal to 29-3 inches of mercury, and have corrected the 

 figures given in columns (e) and (/) in accordance with this 

 assumption. 



The figures given are open to further correction, while 

 some are but approximations. 



In case (3), if the relative humidity be calculated by 

 multiplying the factor 100/ F into the elastic force of vapour 

 at the dew point the result, 0-2882x100^ = 290-8 = 83-8, is 



