Jan. 29, 1885] 



NA TURE 



293 



filled up with deposits from other vials. In shallow places 

 Sphagnum, Utricularia, Myriophyllum, or other finely 

 cut-leaved water-plants should be lifted in the hand, 

 and the water drained or squeezed from them into 

 a tin can, to be subsequently treated in the same 

 way. A few drops of carbolic acid in each vial, 

 just enough to make its presence perceptible, will 

 preserve the contents for months, and even years, 

 from deterioration ; the chlorophyll may fade, but this, 

 in the case of Desmids, is of little importance ; neverthe- 

 less, when practicable, always examine the materials 

 when fresh. When dried on paper for the herbarium, 

 the specimen can still, after being moistened with water, 

 be examined under the microscope, but not with the best 

 results, since the drying up is apt to collapse or otherwise 

 distort the cells. 



The collector will not know the value of his find until 

 it has been brought, drop by drop, under the microscope ; 

 and out of the entire mass he may discern nothing to re- 

 ward his labours. This, however, should not discourage 

 him, as one or two failures are to be expected before 

 meeting with an adequate reward. Sketches ought to be 

 made, which should, of course, be very exact ; and for 

 this purpose the microscope should be provided with an 

 eye-piece micrometer. It is so difficult to separate 

 Desmids from their accompanying foreign matters, that 

 it is seldom amateurs can mount them satisfactorily on 

 slides ; and this method of preserving specimens cannot 

 therefore be recommended. 



RELA TIVE FREQUENCY OF STORMS IN THE 



NORTHERN HEMISPHERE ' 

 "THE portion of the northern hemisphere selected by 



*- the Signal Office of the United States for this dis- 

 cussion is necessarily that part for which the data re- 

 quired are available, and it may be considered as compris- 

 ing a broad belt of from 30^ to 40 of latitude in width, 

 extending from the Pacific sea-board of America, through 

 the United States, Canada, the Atlantic, and Europe, with 

 the North of Africa, eastward into Western Siberia. It 

 thus embraces some of the more important regions of the 

 globe, including the great routes of commerce across the 

 Atlantic. The thirteen charts, which show graphically 

 the relative storm frequency for each month and for the 

 year, have been constructed from data referring to 134 

 months in all, extending from 1863 to 1883. Of the 

 storms which occurred in this extensive region from 

 January 1876 to August 1881, the history of 2730 is 

 briefly summarised. Of these 413 began and ended in 

 America ; 589 began in America and ended in the 

 Atlantic ; 190 began in America and crossed the Atlantic ; 

 326 began and ended in the Atlantic ; 655 began in the 

 Atlantic and ended in Europe ; 491 began and ended in 

 Europe ; and 66 began in America and crossed the 

 Atlantic and Europe. The important bearing of these 

 facts on the telegraphing from America of forecasts of 

 storms about to strike the coasts of Europe scarcely needs 

 to be referred to further than to remark how essential it 

 is for the usefulness of such a service that it be placed in 

 the hands of some competent and responsible central 

 authority in the United States, as was suggested by us in 

 1879 (Nature, vol. xx. p. 359), and which, we believe, 

 has been carried out. 



The chart for the year shows that the region where 

 storms occur with greatest frequency is a long belt in 

 America of about 200 miles in width, extending from the 

 head waters of the Red River, about gs° W. long., east- 

 wards through the Great Lakes to the mouth of the St. 

 Lawrence, about 70 W. long. Surrounding this is a more 



' Charts of Relative Storm Frequency for a Portion of the Northern 

 Hemisphere. Prepared, under the direction of General W. B. Hazen, Chief 

 Signal Officer of the Army, by John P. Finley, Sergeant. Signal Corps, 

 U S.A. Washington : Signal Office, 1884.) 



extensive region where the number of storms, though 

 not so large, is still a good way above the average ; and 

 again, surrounding this latter, is a still wider region, 

 stretching from 105° W. long, eastward through the 

 States and Canada, and through the Atlantic as far as 

 20 long. W. This is one of the most important regions 

 of the globe as regards storms or cyclones. The excessive 

 frequency of storms is probably due to a prevalence, 

 during a large portion of the year, of the south-east trades, 

 with a continuation of easterly and southerly winds into 

 and through the Caribbean Sea and Gulf of Mexico, by 

 which, from the superabundant vapour thus poured north- 

 ward and eastward over the Linked States by upper and 

 lower currents, frequent storms are originated. 



Another region of considerable storm frequency extends 

 from the south of Greenland, through Iceland and Faro, 

 to the north of Sweden. Over this region it may be 

 assumed that a more extended and exhaustive discussion 

 of the storms occurring there than it has been possible to 

 make, will reveal a greater frequency than is indicated on 

 the chart, a supposition rendered highly probable by 

 the frequent and extensive fluctuations of the barome- 

 ter which occur in Iceland during at least three of the 

 four seasons of the year. Of great interest is the less 

 frequency of storms in the Spanish Peninsula and north- 

 eastwards, through Central Europe, as far as Berlin ; 

 and the increased frequency to the southward over the 

 northern half of the Mediterranean and the Black Sea, 

 pointing to the important role played in the storms of 

 that region by the evaporation from these seas. 



This is substantially the distribution of frequency dur- 

 ing the colder months of the year, when the larger number 

 of storms occur. In the spring and summer months the 

 distribution is materially altered. Thus, in April the 

 regions of greater frequency extend further to southward 

 in the United States and the Atlantic. It is in Europe, 

 however, where this southing of the tracks of cyclones is 

 most decidedly marked. At this season a broad patch is 

 seen to overspread Ireland and England, and extend 

 thence southward over the north of Spain, and then east- 

 wards over nearly the whole of the south slope of Europe 

 to near the Caspian Sea. As directly connected with the 

 greater prevalence in spring of cyclones in Southern 

 Europe are the east winds, which acquire at this same 

 season their greatest virulence over the north-western 

 part of the Continent. In summer, on the other hand, 

 the coloured patches marking the regions of greater 

 storm-frequency lie further to the northward than at any 

 other season. Thus, in August, immediately to the north 

 of 50 N. lat., there is an extensive region of greater 

 storm-frequency, of about 900 miles in breadth, extending 

 from about 45° W. long, to eastward as far as St. Peters- 

 burg. In this season the south of Europe is practically 

 rainless, and storms are of extremely rare occurrence. 



From the charts, the tracks usually taken by storms in 

 different parts of the wide region under review cannot be 

 ascertained, but can only be guessed at inferentially. It 

 would be a great improvement if, in subsequent issues of 

 the paper, these tracks were entered on the charts. This 

 was done in 1882 in the "Physical Atlas of the Atlantic 

 Ocean," prepared under the direction of Dr. Neumayer, 

 of Hamburg. It was there shown from centres of the 

 most frequent occurrence of low barometers, that to the 

 west of the Mississippi is the region where most of the 

 L'nited States storms originate ; that many of the Atlantic 

 storms have their origin in the Gulf of St. Lawrence ; 

 and that the storms of North-Western Europe chiefly 

 originate in mid-Atlantic and to the south-west of Ice- 

 land. The centres of low pressure also pointed to a 

 retardation in the onward course of storms on advancing 

 on large masses of land, as happens when storms ap- 

 proach the south of Greenland, the south of the British 

 Islands, Denmark, and the Lofoten Isles. Of all storm- 

 tracks approximately known in the northern hemisphere 



