236 



NA TURE 



[January 9, 1902 



almost wholly to, the western and north-western portions of 

 great storms and blizzards. 



(2) That there seems to be a law of general distribution of 

 the different forms, the columnar to one, the tabular and 

 granular to others, with many varieties associated together in 

 other portions of such great storms. 



(3) That this distribution is, with few exceptions, constant, 

 that iis, the same in nearly all storms. 



Sufficient data has not as yet been collected to demonstrate 

 beyond all doubt the fact that this law applies to all forms of 

 crystals and to all .storms alike. 



Passing on to the variation in form of those crystals de- 

 posited by local storms, as compared with those of general 

 storms, we find that these are very marked, except during 

 intense cold. 



The local storm types and those precipitated from low, de- 

 tached clouds usually consist of large, frail, branching, tabular 

 forms, devoid of a solid tabular nucleus (see No. lo), or of 

 heavy granular varieties, similar one to the other, each ac- 

 cording to its class. On the other hand, those deposited by 

 general storms are usually more diversified in form and more 

 complex in structure, the snowfall often consisting of two or 

 more varieties associated together. The larger and more per- 

 fect columnar prisms, columnar forms possessing tabular out- 

 growths at one or both ends (which we might call doublets), 

 truncated triangular forms (see Nos. 2 and 3), and solid tabular 

 forms, the latter often possessing wonderfully beautiful and 

 complex interior designs (as in No. i), are common only to 

 general storms. Branching tabular and granular forms are 

 common to both general and local storms, but they ordinarily 

 possess solid nuclei if deposited from a general storm (as in 

 Nos. 4-7), whereas the nuclei are generally absent (as in 

 No. 10) if the crystals originated in local storms. During zero 

 weather the crystals of local storms approach much nearer in 

 form to those of general storms, and we find solid tabular 

 forms, branching tabular forms possessing solid hexagonal 

 nuclei and sometimes doublets, among the snowfall. Often 

 during the intense cold .succeeding a blizzard the snowfall will 

 consist wholly of very minute columnar and pyramidal forms, or 

 of both columnar and minute frost-like tabular forms, falling 

 apparently from low, detached nimbus or alto-nimbus clouds, or 

 jven from a sky free, or nearly so, of clouds. 



During relatively mild temperatures each cloud stratum, if 

 alone, there being no other clouds either above or below them, 

 commonly precipitates each its own peculiar type of crystals. 

 Low detached nimbus clouds deposit large, trail, branching 

 tabular forms, similar to No. 10 ; intermediate clouds, smaller, 

 branching tabular forms, possessing solid hexagonal nuclei ; 

 and the high cirro-stratus clouds, small compact columnar and 

 tabular forms. The large cumulus clouds of spring and 

 autumn usually shed large, heavy, pyramidal-shaped granular 

 snow. These granular forms frecjuently, if not invariably, 

 possess nuclei of branching, t.abular forms, and are usually 

 precipitated when the temperature is near or somewhat above 

 the freezing point. 



Of the photomicrographs which accompany this article all, 

 with the exception of No. 10, are those common to and were 

 deposited by great storms. 



Of the other numbers of the series, No. 2 is very rare and 

 unusual, containing as it does eleven triangular divisions 

 within its outlines. Apparently the lines of greatest growth 

 were reversed during one stage of the growth of this strange 

 form, thus differing widely from No. 3, which has outlines 

 somewhat similar. No. 4 possesses a very rare unique nuclear 

 design which is very dillicult to explain by any process of 

 crystallisation of which we know. No. 5 (a souvenir of the 

 great blizzard of March 12, 1888) is very symmetrical, as also is 

 No. 6, of February 15, 1901. No. 7 is, in all but the unimportant 

 outermost points, a marvel of complexity and perfect symmetry. 

 No. 9 is also a marvellously beautiful and symmetrical example 

 of snow architecture. 



Passing to the causes governing the formation of the nucleus, 

 whether it be columnar or tabular, the electrified state of the 

 atmosi)here, whether negative or positive, and perhaps, also, as 

 suggested by Prof. Cleveland Abbe, the presence in greater or 

 less amounts of various gases and vapours in the atmosphere, 

 may all be controlling factors. 



Although much has been already learned about these interest- 

 ing phenomena, yet there still remains much more. Coopera- 

 tion between many observers is essential to carry out this 



NO. 1680, VOL. 65] ' 



work successfully. .Simultaneous observations of Ijie forms and 

 changes the crystals undergo from hour lo hour during our great 

 blizzards should be made by many skilled observers, stationed 

 along a general line extending north and south. These 

 observers must be familiar with the names and approximate 

 heights of the various clouds. This .".tudy should include 

 observations of the kind and appioxiniate height and ilirection 

 of drift of the various clouds, direction and force of the surface 

 wind, temperature of the air, and amount of moisture at the 

 earth's surface ; also its electiic condition, whether negative or 

 positive, and the portion of the storm from which the crystals 

 emanate. 



It is also highly desirable that observations be made to 

 ascertain why the perfect crystals are more common in the 

 western portion of storms, and also why certain portions pro- 

 duce certain types. 



Such a study, supplemented by investigations as to the causes 

 of the formation of the two fundamental types of hoarfrost 

 crystals, would doubtless lead to the discovery of very many ot 

 the mysteries surrounding the origin and history of the 

 wondrously beautiful forms of snow. 



A NEW JOURNAL OF ANATOM\. 

 'X'HE first number of a new scientific magazine — The American 

 Journal of Anatomy (Baltimore, November 1901) — has 

 been received and merits a descriptive notice because, as it has 

 been " founded to collect into one place, and present in a worthy 

 manner, the many researches " of American anatomists, it marks 

 a new departure in scientific journalism, and while its pages are 

 but 98 in number we are informed that future issues will be in 

 quarterly parts of about 1 25 pages each. The interest with 

 which British anatomists regard their branch of science as practised 

 in America has during the last four to five years been heightened 

 by the association, as joint editor of our own long-established 

 Journal of Anatomy ami Pliysiology, of Prof. G. S. Huntington, 

 of the Columbia University at Washington, who is one of the 

 chief promoters of the new journal now under review. The 

 reason of this enhanced interest lies in the fact that his connection 

 with the English publication was marked by the appearance 

 in its p.iges of a paper of a kind to which its readers were 

 unaccustomed. It deals with a series of sections of an early 

 human embryo, and while containing nothing that is new, 

 surpasses all precedent in being illustrated by 1 1 plates of 

 photomicrographic figures which do not portray a single fresh 

 fact and can only be defined as useless. When, further, it is 

 observed that the paper (by an English author) which imme- 

 diately followed this in order of publication was similarly 

 granted 12 plates, where 4 might well have sufficed, there is 

 no wonder that there arose in the minds of the supporters 

 of the journal a misgiving, lest the new association might 

 perhaps lead to disaster. Let it be said, however, that American 

 papers since received have been of a more normal kind. 



This consideration lends interest to the circumstance that 

 in the journal under review one of the five papers submitted 

 monopolises 37 of the 98 pp. which make up the issue, as well 

 as the whole of the 9 plates and 27 of the 42 text figures. 

 It deals with a series of human embryos, as illustrating " The 

 Development of the Limbs, Body-wall, and Back, in Man," and 

 is a joint production by I)rs. C. R. Barden and W. H. Lewis, 

 of Baltimore. It is carefully written, and of the illustrations no 

 praise can be too high. But we deplore the fact that, beyond 

 the more exact determination of the .ictual period at which 

 some of the important constituents of the developing nervous 

 system and parts of the fore- and hind-limbs are first differentiated, 

 there is nothing either recorded or delineated in it which is new. 

 As a chapter for a textbook it would be well-nigh ideal ; but 

 in a journal devoted to records of research and new observations 

 it is out of place and does but hamper the way. 



We wish no disrespect to the authors, for if, .is we assume, 

 in providing the positively magnificent illustrations, they have 

 but availed themselves of the condition set forth in the editorial 

 advertisement, "that the cost of more expensive plates must be 

 borne in part by the authors," we would rather tender them our 

 hearty thanks. While, however, we would thus appraise their 

 enterprise and artistic taste, we are still of opinion that, if our 

 surmise is correct, both these and their enthusiasm have been 

 misapplied ; for if we are to proceed on these lines, the danger, 

 at present obvious enough in all departments of anatomical 



