March 20, 19 19] 



NATURE 



55 



the day self-contained unless we neglect the n.c. (non- 

 cyclic) changes, which are seldom really negligible. 

 These changes should, in any case, be explicitly shown, 

 as they are generally in part of instrumental origin.' 

 The only reference to them at Honolulu seems to 

 be a statement that they have been allowed for in the 

 case of the 5q-day inequalities. At Sitka there is the 

 further statement that, so far as possible, days with 

 large n.c. changes have not been chosen for the loq 

 days. It would be interesting to know how the n.c. 

 corrections were found for the 5q days, and whether 

 they were entirely omitted for the loq days. 



An idea of the size of the n.c. changes on quiet 

 days can usually be derived from the size of the differ- 

 ence between the mean daily values for these days and 

 for all days. From the twenty-four monthly means 

 of 1915 and 1916 given for all days and for the loq 

 days, we find for the mean algebraic excess of the 

 latter class over the former +6oy in H (horizontal 

 force) and +5-47 in V (vertical force) at Sitka, and 

 + 757 in H and -- i-iy in V at Honolulu. In D 

 (declination) the mean difference between the two sets 

 of mean values is only about 005' at both stations. 

 The 5q days give very nearly the same mean daily 

 values as the loq days. We should naturally infer 

 that while the n.c. change may be negligible in D, 

 it is probably by no means negligible in H at either 

 station, or in V at Sitka. Confirmatory evidence is 

 derivable from the highly disturbed days, the charac- 

 teristics of which are usually the direct opposite of 

 those of quiet days. If we take as representing dis- 

 turbance the five days of largest daily range in each 

 month, we find that on the average the monthly mean 

 values derived from these days fall short of the corre- 

 sponding all-day means by 14-47 in H and 15-37 in V 

 at Sitka, and by 9- 17 in H and 2-37 in V at Honolulu. 

 Disturbance in V is unusually large at Sitka, and 

 exceptionally small at Honolulu, which, presumably, 

 <xplains the large difference between the mean values 

 from the 5q and 5d days at the former station, and 

 the small difference at the latter station. If a diurnal 

 inequality were to be derived from the 5d days, as 

 has recently been suggested, the n.c. element would 

 almost certainly be of great importance at Sitka in H 

 and V. 



Another feature v^'anting explanation is that the 

 5q days are shown in the tables as days of 165° W. at 

 Honolulu, and as days of 135° W. at Sitka, whereas 

 thev are reallv 24-hour periods commencing at Green- 

 wich midnight. It is to be hoped that this is only 

 camouflage, just as when values belonging really to 

 i-5h. are entered under the heading 2h., because 

 considerable disturbance is occasionally experienced 

 within less than nine hours of the endl of true inter- 

 national quiet days. Explanation on this point, on the 

 n.c. changes, and on the effect on the ranges of the 

 diurnal inequalities consequent on the change of pro- 

 cedure would be welcome in the next issue of these 

 valuable publications. C. Chree. 



FOREST RESEARCH IN EUROPE. 



A USEFUL account of forest research in Europe 

 by Mr. S. Howard has appeared in the Indian 

 Forester for September last. "Forest research, in 

 many instances, necessitates observations over long 

 periods of time, longer than an individual man's 

 working vears, and over widely separated areas. 

 Some institution is necessary, therefore, to direct 

 methods for the sake of uniformity, and to continue 

 ideas, despite the necessary changes in the research 

 personnel.'^ Germany was the first country to 

 organise research, this movement dating from 1868, 

 when it was proposed at a meeting of prominent 



NO. 2577, VOL. 103] 



foresters at Regensburg that the larger States, Aus- 

 tria, Prussia, and Bavaria, should have independent 

 research institutes. It was finally decided in 1870 that 

 forest research should be properly organised, and that 

 the research institutes in all the States were to be 

 combined with the educational branch — that is to say, 

 the president of the forest college was also to be presi- 

 dent of the research institute. In 1912 all German 

 States of importance had their forestry institutes (in 

 each case combined with the college of forestry) united 

 under the German Forest Research Association, which 

 meets, as a rule, twice a year. The Prussian Research 

 Institute, united with the Forestry College at Ebers- 

 walde, has six branches, dealing with sylviculture, 

 physical chemistry, meteorology, plant physiology, 

 zoology, and mycology. 



Besides the headquarters at Eberswalde, there are 

 numerous experimental plots of trees all over Prussia. 

 These were at first put under the local forest officer, 

 but this proved. a failure; and for more than twenty 

 years all work connected with the plots has been done 

 by the research staff at Eberswalde. The sylvicultural 

 branch compiles yield tables and statistics, and has 

 carried out investigations on exotic trees, root- 

 formation, manures, technical properties of wood, 

 seed tests, etc. It is to be noted that the tests of 

 strengths of wood and the like are done by technical 

 experts at Charlottenburg, and not by the Forest 

 Research Institute. The meteorology branch is 

 especially concerned with experiments on the influence 

 of forests on climate. The plant physiology branch is 

 purely botanical, and takes up subjects like the forma- 

 tion of annual rings, the influence of locality on seeds, 

 the influence of soil factors on trees, etc. The zoolog}' 

 branch is concerned with zoological researches, so far 

 as they concern forests, and with control methods ; 

 the physical chemistry branch with the chemistry of 

 soils, the formation of humus, the formation of pan, 

 etc. ; and the mycology branch with mycology in its 

 relation to forestry and control methods. 



Forest research was organised in France in 1882, 

 but was hopelessly carried out. Experiments were 

 begun, but were usually badly organised and badly 

 performed, and ceased absolutely between 1896 and 

 1902. The research work proposed in France was to 

 cost 1200I. per annum. Germany has actually been 

 spending 6oooi., Switzerland 2000Z., and Sweden, 

 roughly, 88oi!. a year. 



Mr. Howard gives also an account of the Inter- 

 national Forest Research Association, which met at 

 Mariabrunn in 1893 and 1903, at Brunswick in 1896, 

 at Zurich in 1900, and at Brussels in 1910.^ Most 

 European countries are members, but France is not, 

 or, if she is, has taken no active part. The usefulness 

 of these international meetings was undoubted. For 

 example, it was soon found essential to have some 

 classification of thinnings, if results were to be com- 

 parable. The Prussian classification was adopted in 

 1503. 



THE CONSERVATION OF OUR CEREAL 

 ^ RESERVES.^ 



THE dangers to which grain stored under ordinary 

 conditions is exposed may be classified under 

 four heads :-^i) The attacks' of rats and mice, 

 (2) those of insects and mites, (3) those of moulds 

 and bacteria, and (4) the process known as 

 " heating." The amount of damage due to rats and 

 mice is, no doubt, enormous, but might be avoided 

 by any rational system of storage, and is a matter 



' Abstract of a leaure delivered at Kins's ColIe(re, London, on M»rch 121 

 under the auKoices of the Imperial Studies Committee of the Univer&ity 

 Ixjndon, by Prof. Arthur Dendy, F.F.S. 



