342 



NATURE 



\^Atigust 12, 1880 



greatest possible number, if from one locality and horizon, 

 be included together. 



There are not wanting altogether, however, indications 

 of other species, and among tbem G. rigida, G. rotula, 

 and G. niicroincra seem to be distinct, but the great 

 majority are simply pectinato-pinnatifid, and possess no 

 really distinctive specific characters. In addition to this, 

 fourteen species from one locality and horizon appear 

 a very unlikely number to have existed together, for 

 although the plants are sociable and grow massed toge- 

 ther, but few species are ever met with living together in 

 the same vicinity. The whole of America, which is the 

 richest continent in species, contains but nine, the varied 

 lands grouped as the Malayan region but seven, New 

 Zealand five, Australia four, &c. ; the total number recog- 

 nised by Hooker in the "Synopsis Filicum " being but 

 twenty-three. The greatest number growing in a re- 

 stricted area is in North Caledonia, where there are 

 four ; but I am not aware whether these are actually 

 associated together. 



These Gleichenias are repeatedly alluded to by Prof. 

 Heer as indicating a tropical nature for the Arctic creta- 

 ceous flora, but so far as their presence goes, they by no 

 means imply that a high temperature prevailed. Although 

 no Gleichenia now ranges into high northern regions, 

 they flourish south in the rigorous climates of the Magel- 

 lan and Falkland Isles, S. lat. 53', which have an 

 isotherm of less than 45°, and are also found on the high 

 mountains of Tasmania and on the Andes at 10,000 feet, 

 which is, according to Humboldt, the level of gentians 

 and near the limit of arborescent vegetation. The group 

 of Gleichenias from the colder regions of South America 

 all resemble each other in much the same degree as 

 those of the Arctic regions did, and all possess small, 

 hard, rigid, pectinato-pinnatifid pinna;. Among them 

 are G. pedalis, G. ayp/ocarpa, and G. quadripartita, all 

 of which, but especially the former, vary considerably, 

 being either long or shortly pectinate. It is a suggestive 

 fact that the existing representative of these Arctic 

 Gleichenias is the only one that still ranges into northern 

 temperate regions, such as China and Japan, while the 

 representative of the English Eocene species is an essen- 

 tially tropical form. 



The Arctic group of Gleichenia appears to have very 

 little affinity with European fossil plants of similar age, 

 except through G. Zippci. Hecr connects one with 

 G. comptoniafolia, from Aix-la-Chapelle, although there 

 is little discoverable resemblance between them. To do 

 so he has to point out a discrepancy between the drawing 

 and the description, and although he had never seen the 

 specimens, prefers to rely on the drawing which Dr. De 

 Bey now disclaims as incorrect. The Aix-la-Chapelle 

 types are really quite different and more varied, and link 

 ihem with our own eocene species. This latter is an 

 essentially tropical type, and completely distinct from 

 either the fossil Arctic group or the existing forms from 

 the cold southern latitudes, since it closely approaches 

 G. dicJtotoiiia, the only type of a well-defined section of 

 the genus, now almost universally distributed over the 

 tropical world. 



The Gleichenias seem first to have appeared in the 

 Jurassic, to have passed away from Europe before the 

 close of the Eocene period, and to be now decidedly 

 characteristic of the southern hemisphere — very few 

 species crossing the equator, although the representative 

 of the fossil Arctic species still extends as far north as 

 Japan. It is obvious that we need not, from their presence, 

 assign a very high mean annual temperature to the older 

 cretaceous period in Greenland. J. S. Gardner 



METEOROLOGY IN JAPAN 

 V^TE have read carefully and with great pleasure the 

 Memoirs of the Science Department of the Uni- 

 versity of Tokio, Japan, vol. iii. Part i., which gives the 



report of the meteorology of Tokio for 1879, by Prof. 

 T. C. Mendenhall. The observations, which are carried 

 on in the west wing of the small observatory attached to 

 the University, were commenced in January 1S79, and 

 this is the first report issued by the Observatory. The 

 instruments are from Negretti and Zambra, and, with the 

 exception of the thermometers, they appear to have been 

 placed in suitable positions. The thermometers are 

 mounted outside the north window of the second floor, 

 and are separated from the observing room by glass 

 doors, which are opened for observation. This position 

 of the thermometer is in several respects objectionable, 

 but particularly as it precludes any comparability, beyond 

 a rough one, between the temperature observations at 

 Tokio and at other stations which are or may be esta- 

 blished in Japan. 



The hours of observation are 7 a.m., 2 and 10 p.m., an 

 arrangement of hours, it may be remarked, which states 

 the mean temperature of the six warmest months of the 

 year about three-fourths of a degree too high, and further 

 does not approximate with the desired closeness to the 

 important diurnal turning-points of the barometric pres- 

 sure. It is however right to add that it is declared 

 desirable to increase the number of the observations to at 

 least five or six during the day as soon as the necessary 

 arrangements can be made, and to institute a series of 

 hourly observations for approximately determining several 

 of the diurnal curves. An arrangement, if possible to be 

 carried out, for the erection of continuously-recording 

 instruments, would be an important gain to Japan 

 meteorology. 



The observations are published in extenso, and are 

 illustrated with great fulness by excellent diagrams, which 

 show in a clear manner the main results of the year's 

 observations, the diagrams being lettered and numbered 

 so as to serve for both the English and the Japanese 

 editions which are issued. 



The mean pressure for the year at 32° and sea level is 

 given at 29952 inches, the monthly maximum, 30'093 

 inches, having occurred in January, and the minimum, 

 29'8o9 inches, in ."August, thus showing a tendency in the 

 atmospheric pressure to be assimilated to the annual 

 march of pressure in the continent adjoining. There 

 having occurred no typhoon during the year, the lowest 

 barometer was only 29'o87 inches, which happened on 

 February 23, and the highest, 30-5 15 inches, on April 21, 

 the range for the year thus being 1-426 inch. The mean 

 diurnal range from 7 a.m. to 2 p.m. is large, being 0059 

 inch for the year, regarding which Prof. Mendenhall 

 remarks that "this same relation exists in each set of 

 monthly means with two exceptions.'' These exceptions 

 are May and September, the ranges for which being, as 

 printed in the means, 0'02S inch and o'oi9 inch. On 

 comparing these ranges with those for the other months, 

 they are at once seen to be physically impossible ; but by 

 averaging the observations themselves for these months 

 these exceptiona'ly low ranges turn out to be due solely 

 to errors of computation. The true range given by the 

 observations for May and September are 0-047 inch for 

 each month. The exceptionally large range for July, viz., 

 0085 inch, is also an error of computation; the true 

 range was onlyo'052 inch, the mean range at Tokio being, 

 as in corresponding latitudes of the Atlantic, less in the 

 summer than in the winter months. 



The lowest temperature for the year was 24°T on 

 Janmry 2 and 7, and the highest 93°-o on August 15. The 

 temperature fell to or below freezing (32") on 46 days, 27 

 of these days being in January, and rose to or above 90° 'o 

 on 12 days, 7 of these days being in July and 5 in August. 

 The mean annual temperature deduced from the 7 a.m., 

 2 and 10 p.m. observations was 58°-5, and from the maxi- 

 mum and minimum observations sS'-o, the higher tem- 

 perature of the former being due to the 7 a.m. observations. 

 If this were changed to 6 a.m. the hours of observatioa 



