or subgroups of the characters in a definite order corresponding to the real (though in detail undeterminable) quantitative scale in the character or attribute itself. The order of the classes appeared to be the important thing, and consequently the method was assumed to be limited to such attributes as could be arranged in a definite scale order. In recent work, however, by varying the order of the classes Pearson has found that so far as the value of the correlation coefficient is concerned this group order has practically no influence. For the new conception of correlation which arose from a consideration of this fact Pearson proposes the term contingency.

As a measure of the contingency of any classification of characters, it is proposed to use some measure of the total deviation of the classification from independent probability.' The practical method of making such a measure Pearson develops in the following

way.

"Let A be any attribute or character and let it be classified into the groups A,, A., **, A, and let the total number of individuals examined be N, and let the numbers which fall into these groups be n1, n, ́ ́ ́, n ̧, respectively. Then the probability of an individual falling into one or the other of these groups is given by n/N, n/N,, n/N, respectively. Now suppose the same population to be classified by another attribute into the groups B1, B,,, B, and the group frequencies of the N individuals to be m,, m,,,m,, respectively. The probability of an individual falling into these groups will be respectively m,/N, m/N, m/N,, m/N. Accordingly the number of combinations of B, with A" to be expected on the theory of independent probability if N pairs of attributes are examined is

uv

B..

v

in the occurrence of the groups A Clearly the total deviation of the whole classification system from independent probability must be some function of the n quantities for the whole table." The value of any function of these quantities will clearly be independent of the order of classification. The following functions of the n. 19 quantities were chosen for practical use. (a) 1-P; the contingency grade, where P is determined from x' by the use of Elderton's tables.* The quantity x' is a measure of the deviation of the observed results from independent probability, depending on the quantities as shown by the equation

uv

u v

น

This function

02

is called the first coefficient of contingency and is denoted by C1.

The analysis of the relation of function in the case of normal correlation leads to the practical result that the value of r may be obtained if is given, which, of course, is the case, the latter function being obtained from the observations. A table and plotted curve from which values of r correct to two places may be read off directly, are given. If the coefficient so obtained from be designated as C, the second coefficient of contingency, we have as a limiting case

After considering the subject of multiple contingency and its relation to multiple normal correlation the author proceeds to give some illustrative examples showing something of the sort of problems to which the method may be applied, and also how it is to be used in practise. The examples include (a) the correlation between father and son in respect to stature, (b) color inheritance in greyhounds, (c) fraternal resemblance in hair color in man, and (d) the correlation between father and son in respect to occupation or profession.

The net results brought out by the analysis and confirmed by the numerical illustrations may best be stated in the author's own words:

"With normal frequency distributions both contingency coefficients pass with sufficiently fine grouping into the well-known correlation coefficient. Since, however, the contingency

is independent of the order of grouping, we conclude that, when we are dealing with alternative and exclusive sub-attributes, we need not insist on the importance of any particular order or scale for the arrangement of the subgroups. This conception can be extended from normal correlation to any distribution with linear regression; small changes (i. e., such that the sum of their squares may be neglected as compared with the squares of mean or standard deviation) may be made in the order of grouping without affecting the correlation coefficient." These results "are not so fruitful for practical working as might at first sight appear, for they depend in practise on the legitimacy of replacing finite integrals by sums over a series of varying areas, where no quadrature formula is available. If we, to meet the difficulty, make a very great number of small classes, the calculation, especially of the mean square contingency, becomes excessively laborious. Further, since in observation individuals go by units, casual individuals, which may fairly represent the frequency of a considerable area, will be found on some one or other isolated small area, and thus increase out of all proportion the contingency. The like difficulty occurs when we deal with outlying individuals in the case of frequency curves, only it is immensely exaggerated in the case of frequency surfaces. It is thus not desirable in actual practise to take too many or too fine subgroupings. found, under these conditions, that the correlation coefficient as determined by the product moment or fourfold division methods is approximated to more closely in the case of the contingency coefficient found from mean square contingency than in the case of that found from mean contingency. Probably 16 to 25 contingency subgroups will give fairly good results in the case of mean square contingency, but for each particular type of investigation it appears desirable to check the number of groups proper for the purpose by comparing with the results of test fourfold

It is

division correlations. Under such conditions it appears likely that very steady and consistent results will be obtained from mean square contingency."

In the calculation of contingency coefficients the present writer has found that the following procedure saves much time and labor. The value of the independent probability v for each compartment of the table is obtained by the use of a Thacher calculating instrument (Keuffel and Esser). With this instru

Vur

ν

ment one can read directly to four or five figures the values of any expression which can be put into the form ax/b, where a and b are constants and x is a variable. Since for any compartment equals (nm)/N for that compartment, it is evident that by taking either n or m, as the constant, it will only be necessary to make as many settings of the instrument as there are rows or columns in the table. Having obtained the v quantities, the sub-contingencies (nv) may be written down directly, squared from Barlow's tables, and divided by v with an arithmometer or with Zimmermann's or Crelle's multiplication tables. The remainder of the calculations necessary to obtain the mean square contingency and the whole of the calculations for the mean contingency, and their respective coefficients are, of course, easily performed. Proceeding in this way, the calculation of contingency coefficients, even though several experimental groupings are made, has been found to take but comparatively little time.

The noteworthy features of this method of contingency are found in that it, in the first place, broadens and illumines the whole theory of correlation, and in the second place, brings within the range of biometrical investigation a large series of problems to which it has hitherto been impossible to apply exact methods. One can but feel that this memoir, like so many of the others which have preceded it in the series, marks a definite and fundamental step in advance in the steady progress of the science of biometry.

lium' was gone into quite carefully by Professor F. W. Clarke and also by the committee appointed by the American Association on the Spelling and Pronunciation of Chemical Terms, and the conclusion was arrived at that the name 'glucinum' should be used on the ground of priority. In SCIENCE for December 9 Dr. Charles Lathrop Parsons has stated his grounds for preferring the name 'beryllium.' Dr. Parsons is, thanks to his bibliographical work on the element in question, thoroughly informed in its literature, but the arguments adduced by him would seem to lead to a conclusion diametrically opposed to that which he has drawn.

At

It was obviously the privilege of Vauquelin, the discoverer of the element, or rather its oxid, to name it. This he never did, but contented himself by speaking of it at first as 'la terre du Béril,' that is, the earth in beryl. the close of Vauquelin's first paper the editors of the Annales added a note signed 'Redacteur' in which they propose the name 'glucine.' It was of course well known that Guyton and Fourcroy were the editors. Vauquelin's second paper in the Annales was evidently prepared at the same time as the first, or at least before the second was in print. In his third paper, some weeks later, as Dr. Parsons admits, Vauquelin actually adopted the term 'glucine,' prefacing its use with 'on a donné le nom de glucine.' The paper in the Journal des Mines was apparently prepared at the same time as the first two papers in the Annales and before the appearance of the suggestion of Guyton and Foureroy, but at its close occurs the note which Dr. Parsons has quoted. In this he states that Guyton and Fourcroy have advised him to call the new earthglucine' and while he evidently does not think the name the best that could have been chosen, he clearly acquiesces in the suggestion of the two great authorities and says 'Cette denomination sera assez significante pour aide le mémoire.' Finally, as seen above, in his third paper, he adopts the name. As far as priority goes, the argument in favor of 'beryllium' would seem to be that probably Vauquelin would have given the earth some other name had he ventured to dissent from

Guyton's authority, and it is probable that he would have liked to name it 'beryllia.' All of which may be quite true, but actually he did not do it.

As regards the German use of 'Berylerde ' it was merely at first the natural translation of Vauquelin's expression 'la terre du Béril,' which, as we have seen, he used in no denominative sense. If the generally accepted rules of priority have any weight 'glucinum' is the only term to be used for the element.

As regards usage, the case is hardly quite as bad as Dr. Parsons seems to think, since the index to the Journal of the Chemical Society (London) for 1903 gives 'Beryllium, see glucinum.' With French, English and Americans using 'glucinum,' we can afford to let the German journals cling to 'beryllium' a little while longer.

Incidentally, what shall we do when the Germans insist on kalzium, kolumbium, karolinum, zerium and zesium, or will it be kæsium? JAS. LEWIS HOWE.

WASHINGTON AND LEE UNIVERSITY,

December 12, 1904.

BOTANICAL NOTES.

THE STUDY OF FIBERS.

THE book (The Textile Fibers, their Physical, Microscopical and Chemical Properties') prepared by Dr. J. M. Mathews, and recently published by John Wiley, should make the study of textile fibers somewhat easier by students and practical operators. It covers nearly three hundred pages of neatly printed text, illustrated by sixty-nine cuts, in which the author has presented the whole matter in a most helpful way. There is first a useful classification of fibers, followed by descriptions and discussions of those which enter into fabrics. Some of these fibers are, of course, of animal origin, as wool, hair and silk, and to these are given about ninety pages. The remainder of the book is devoted almost wholly to plant fibers, and here the treatment is especially clear and helpful. The origin. varieties, physical and chemical properties of cotton, and mercerized cotton, are discussed in as many chapters. Linen is given another chapter, while jute, ramie, hemp and several

other fibers of minor importance are disposed of in another chapter. An interesting chapter for the general reader is the one on artificial silks, the processes for the production of which 'have been attended with a considerable degree of success.' It is said that artificial silk has become a commercial article, and is used in considerable quantity by the textile trade.' Of these artificial silks there are four general kinds, viz:

1. Pyrozylin silks, made from a solution of gun cotton in a mixture of alcohol and ether. 2. Fibers made from a solution of cellulose in ammoniacal copper oxide or chloride of zinc.

3. Viscose silk, made from a solution of cellulose thiocarbonate.

4. Gelatin silk, made from filaments of gelatin rendered insoluble by treatment with formaldehyde.

Most of the artificial silk is of the first variety, the manufacture of which is carried on in England, Germany, France and Switzerland. "The fibers are formed by forcing the ether-alcohol solution of pyroxylin through glass capillary tubes, and winding them on frames. As the solution is very viscous it requires a pressure of forty-five atmospheres to discharge it through the capillary openings."

A STUDY OF COMPARATIVE EMBRYOLOGY.

THE Comparative embryology of the Cucurbitaceae (Gourd Family) has been studied by Dr. J. E. Kirkwood, the results of which appear in the Bulletin of the New York Botanical Garden (No. 11, 1904). After an instructive historical introduction, the organogeny of representatives of the five tribes (Fevilleae, Melothrieae, Cucurbiteae, Sicyoideae, and Cyclanthereae) is summarily described, and this is followed by a quite particular examination of the embryo-sac in sixteen genera distributed among the five tribes. Twelve fine plates of 166 figures add much to the value of this portion of the paper. In a closing discussion the author finally concludes that in most points the differences between the Cucurbitaceae, and other sympetalous families are more striking than the similarities.' The paper closes with a bibli

ography including 89 titles. It constitutes a valuable addition to our knowledge of the embryology of a family whose place in the system of plants is still in doubt.

A HELPFUL BULLETIN.

THE office of experiment stations of the United States Department of Agriculture has issued a bulletin (No. 2) consisting of an outline of a lecture on 'Potato Diseases and their Treatment,' for the use of farmers' institute lecturers. It was prepared by F. C. Stewart and H. J. Eustace, of the New York Experiment Station. It contains summaries of our knowledge of the most important diseases which affect the potato in the United States. The descriptions are given in nontechnical language, and ought to convince every botanist of the possibility of treating quite difficult subjects in plain English. Following the description of diseases, is an admirable chapter on spraying and other preventive measures. A very useful bibliography is added in an appendix.

CHARLES E. BESSEY. THE UNIVERSITY OF NEBRASKA.

THE NOBEL PRIZES.

IN a cablegram from Stockholm to the London Times, dated December 10, further details are given in regard to the Nobel prizes.

The prize for physics has been awarded to Lord Rayleigh, professor of natural philosophy at the Royal Institute. The chemistry prize is conferred upon Sir William Ramsay, professor of chemistry at University College. M. Pavloff, professor at the Military Academy of Medicine at St. Petersburg, receives the prize for physiology and medicine. The literature prize is divided between M. Mistral, the Provençal poet, and Don Jose Echegaray, the Spanish dramatist. The peace prize has been awarded to the Institute of International Law.

The distribution of the Nobel prizes took place in the great hall of the Academy of Music at Stockholm in the presence of King Oscar. Lord Rayleigh, Professor Ramsay and M. Pavloff received their prizes, together with diplomas and gold medals, in person

from his Majesty, while the prizes awarded to M. Mistral and Don Jose Echegaray, who were unable to be present, were handed to the French and Spanish ministers respectively. The sum of money attaching to each prize amounts to 140,858 kroner (about $39,000). The Nobel peace prize will be presented by the Norwegian Storthing at Christiania.

The distribution of the prizes was followed by a banquet at the Grand Hotel. Covers were laid for 190 guests, the company including the Crown Prince, Prince and Princess Charles, Lord and Lady Rayleigh, Sir William and Lady Ramsay and M. and Mme. Pavloff. Count Mörner, speaking in German, proposed the health of M. Pavloff; Professor Petterson, in English, proposed the health of Sir William Ramsay; and Professor Hasselberg, in Latin, that of Lord Rayleigh.

SCIENTIFIC NOTES AND NEWS.

AT the meeting of the American Association for the Advancement of Science held at Philadelphia last week, Professor C. M. Woodward, of Washington University, was elected president for the New Orleans meeting.

Ar the recent Philadelphia meeting of the American Society of Naturalists, Professor William James, of Harvard University, was elected president. Professor Chas. B. Davenport, of the Cold Spring Laboratory of Experimental Evolution of the Carnegie Institution, and Professor J. M. Coulter, of the University of Chicago, were elected vice-presidents, and Professor W. E. Castle, of Harvard University, secretary.

PROFESSOR MARY WHITON Calkins of Wellesley College, has been elected president and Mr. Wm. Harper Davis, of Lehigh University, secretary, of the American Psychological Association.

PROFESSOR JOHN DEWEY, of Columbia University, has been elected president of the American Philosophical Association.

PROFESSOR S. W. BURNHAM, astronomer at the Yerkes Observatory, has been awarded the Lalande gold medal of the French Academy of Sciences for his researches in astronomy.

PROFESSOR SVANTE ARRHENIUS has been made head of a laboratory for physical chem