accuracy and elegance; therefore linguistic training has had great importance in the idea of cultivation. The conditions of the educated world have, however, changed so profoundly since the revival of learning in Italy that our inherited ideas concerning training in language and literature have required large modifications.

In the year 1400 it might have been said with truth that there was but one language of the scholars, the Latin, and but two great literatures, the Hebrew and the Greek. Since that time, however, other great literatures have arisen, the Italian, Spanish, French, German, and, above all, the English, which has become incomparably the most extensive and various and the noblest of literatures.

Under these circumstances it is impossible to maintain that a knowledge of any particular literature is indispensable to culture. When we ask ourselves why a knowledge of literature seems indispensable to the ordinary idea of cultivation, we find no answer except this-that in literature are portrayed all human passions, desires and aspirations, and that acquaintance with these human feelings and with the means of portraying them seems to us essential to culture. The linguistic and literary element in cultivation therefore abides, but has become vastly broader than formerly, so broad, indeed, that selection among its various fields is forced upon every educated youth.

The store of knowledge. The next great element in cultivation to which I ask your attention is acquaintance with some parts of the store of knowledge which humanity in its progress from barbarism has acquired and laid up. This is the prodigious store of recorded, rationalized and systematized discoveries, experiences and ideas-the store which we teachers try to pass on to the rising generation.

The capacity to assimilate this store and.

improve it in each successive generation is the distinction of the human race over other animals. It is too vast for any man to master, though he had a hundred lives instead of one; and its growth in the nineteenth century was greater than in all the thirty preceding centuries put together. In the eighteenth century a diligent student with strong memory and quick powers of apprehension need not have despaired of mastering a large fraction of this store of knowledge. Long before the end of the nineteenth century such a task had become impossible.

Culture, therefore, can no longer imply a knowledge of everything-not even little knowledge of everything. It must be content with general knowledge of some things, and a real mastery of some small portion of the human store. Here is a profound modification of the idea of cultivation which the nineteenth century has brought about. What portion or portions of the infinite human store are most proper to the cultivated man? The answer must be-those which enable him, with his individual personal qualities, to deal best and sympathize best with nature and with other human beings.

It is here that the passion for service must fuse with the passion for knowledge. We have learned from nineteenth century experience that there is no field of real knowledge which may not suddenly prove contributory in a high degree to human happiness and the progress of civilization, and therefore acceptable as a worthy element in the truest culture.

Imagination. The only other element in cultivation which time will permit me to treat is the training of the constructive imagination. The imagination is the greatest of human powers, no matter in what field it works-in art or literature, in mechanical invention, in science, government, commerce or religion, and the

training of the imagination is, therefore, far the most important part of education.

I use the term constructive imagination, because that implies the creation or building of a new thing. The sculptor, for example, imagines or conceives the perfect form of a child ten years of age; he has never seen such a thing, for a child perfect in form is never produced; he has seen in different children the elements of perfection, here one and there another. In his imagination, he combines these elements of the perfect form, which he has only seen separated, and from this picture in his mind he carves the stone and in the execution invariably loses his ideal-that is, falls short of it or fails to express it. Constructive imagination is the great power of the poet, as well as of the artist, and the nineteenth century has convinced us that it is also the great power of the man of science, the investigator and the natural philosopher. The educated world needs to recognize the new varieties of constructive imagination.

Zola, in 'La bête humaine,' contrives that ten persons, all connected with the railroad from Paris to Havre, shall be either murderers or murdered, or both, within eighteen months; and he adds two railroad slaughters criminally procured. The conditions of time and place are ingeniously imagined, and no detail is omitted which can heighten the effect of this homicidal fiction.

Contrast this kind of constructive imagination with the kind which conceived the great wells sunk in the solid rock below Niagara that contain the turbines that drive the dynamos, that generate the electric force that turns thousands of wheels and lights thousands of lamps over hundreds of square miles of adjoining territory; or with the kind which conceives the sending of human thoughts across 3000 miles of stormy sea instantaneously on nothing more substantial than ethereal waves.

There is going to be room in the hearts of twentieth century men for a high admiration of these kinds of imagination as well as for that of the poet, artist or dramatist.

It is one lesson of the nineteenth century, then, that in every field of human knowledge the knowledge the constructive imagination finds play-in literature, in history, in theology, in anthropology, and in the whole field of physical and biological research.

That great century has taught us that, on the whole, the scientific imagination is quite as productive for human service as the literary or poetic imagination. The imagination of Darwin or Pasteur, for example, is as high and productive a form of imagination as that of Dante, of Goethe, or even Shakespeare, if we regard the human uses which result from the exercise of imaginative powers, and mean by human uses not meat and drink, clothes and shelter, but the satisfaction of mental and spiritual needs.

It results from this brief survey that the elements and means of cultivation are much more numerous than they used to be; so that it is not wise to say of any one acquisition or faculty-with it cultivation becomes possible, without it impossible.

The one acquisition may be immense, and yet cultivation may not have been attained. We have met artists who were rude and uncouth, yet possessed a high degree of technical skill and strong powers of imagination. We have seen philanthropists and statesmen whose minds have played on great causes and great affairs, and yet who lacked an accurate use of their mother tongue, and had no historical perspective or background of historical knowledge. We must not expect systematic education to produce multitudes of highly cultivated and symmetrically developed persons; the multitudinous product will always be imperfect, just as there are no perfect trees, animals, flowers or crystals.

Let us as teachers accept no single element or variety of culture as the one essential; let us remember that the best fruits of real culture are an open mind, broad sympathies and respect for all the diverse achievements of the human intellect at whatever stage of development they may be to-day-the stage of fresh discovery, or bold exploration, or complete conquest. The moral elements of the new education are so strong that the new forms of culture are likely to prove themselves quite as productive of morality, high-mindedness and idealism as the old.

CHAS. W. ELIOT.

SCIENTIFIC BOOKS.

West Indian Madreporarian Polyps. By J. E. DUERDEN. Memoirs of the National Academy of Sciences, Vol. VIII. 1902.

It may seem strange that notwithstanding the thorough study that has been devoted to the skeleton of the corals our knowledge of their soft parts has been exceedingly limited until recent years. It must be remembered, however, that interest in the anatomy of the nearly related Actiniaria was not really awakened until the publication of Richard Hertwig's report on the Challenger collection in 1882, and the technical difficulties in the way of extended anatomical study of the coral may well be advanced as an excuse for its neglect.

In the same year that Hertwig's report appeared, however, von Koch laid the foundation for a proper appreciation of the significance of the soft parts of the corals by demonstrating the ectodermal nature of the corallum, and since that date valuable contributions to the anatomy of the Madreporarian polyps have been made by von Koch himself and by Bourne, Fowler and von Heider. The total number of forms studied has, however, remained comparatively small, and although enough information was gained to demonstrate a close similarity of the Madrepores to the Hexactiniæ, yet there was a lack of suffi

cient data upon which general conclusions could be based. A systematic study of a large number of forms was needed, and this need has recently been supplied by Dr. J. E. Duerden in his paper on the West Indian corals, a paper destined to stand as a landmark in our knowledge of Madreporarian morphology equal in importance to that established by von Koch.

Duerden has made a thorough study of the morphology of no less than twenty-six species of corals belonging to nineteen different genera, and, when the difficulties in the way of such work are properly appreciated, nothing but admiration can be expressed for the patience, perseverance and thoroughness evidenced in every page of his work. It is monographic in its nature, considering in detail the structure, histology and development of the coral polyps as a group, and concluding with full descriptions of the special morphology of the various forms studied.

Dr. Duerden gives good reason for believing that all corals are fundamentally hexamerous, the corallum septa making their appearance symmetrically in embryos already provided with the six pairs of primary mesenteries. In some species the hexamerism becomes much obscured in later stages, while in others it is more or less distinctly preserved; and it has been possible to correlate these differences with the mode of non-sexual reproduction followed by the species. Two principal methods of non-sexual reproduction are recognizable, namely, gemmation and fission. In the former method the mesenteries of the new individual are formed de novo and repeat the embryological development, and consequently the hexamerism of those of the parent, while in the latter method half the mesenteries of the parent pass directly to each descendant whose growth processes are limited to an attempt to reproduce the lacking parts, a second fission frequently supervening before the attempt is carried to completion. In the polyps produced by gemmation the mesenteries present the usual hexamerous arrangement, two pairs of directives and at least four additional pairs arranged symmetrically to the

directives being present. But in fissiparous forms, since there is no tendency to reproduce directives in the regenerative growth which succeeds the division, all of the polyps of a colony, with the exception of two, will lack directives and will show little regularity in the arrangement of the mesenterial pairs.

What might be regarded as a third mode of non-sexual reproduction has been observed in the perforate corals Madrepora and Porites and has been aptly termed by Duerden fissiparous gemmation, since the mesenteries of any one polyp are partly derived directly from the parent and are partly new formations, the process in this respect resembling ordinary fission, while it also resembles gemmation in that the original hexamerous arrangement of the mesenteries and the typical number of directives are retained as a result of growth processes which precede the fission.

The careful study of the madrepores, however, has not yielded such important taxonomic results as might have been expected; their soft parts do not present as much variety as do those of the actinians. But by extending his observations over so great a number of forms Duerden has been able to establish as fundamental certain facts in the morphology of the corals which throw some light upon their position among the Anthozoa. The fact that the corallum appears only after the development of the first cycle of mesenteries seems to warrant the conclusion that the corals are derived from non-coralligenous hexamerous forms. In other words, it indicates that the Hexactiniæ, Zoanther and Madreporaria are all traceable to a common hexamerous ancestor, and that after the differentiation of the Zoanthem the Madrepores and Hexactinians continued together for a time and have for a fundamental distinction only the development or non-development of a corallum. The Madreporaria are merely Hexactinia which secrete a corallum. This is by no means a novel view of the relationship of these two groups, but it is one that is emphasized by Dr. Duerden's careful and interesting observations.

But the question whether the derivation of

the Madreporaria from the Hexactinia is mono- or polyphyletic still lacks a decisive answer. The uniformity of structure shown by the madreporarian polyp seems to argue for a monophyletic origin, although it by no means excludes the other possibility. It is exceedingly interesting to note that of all the actinians, those which approach nearest to the corals in structure are, as Duerden himself has elsewhere pointed out, such forms as Actinotryx and Ricordea, forms, that is to say, belonging to the stichodactyline group of actinians, having more than one tentacle arising from certain of the endocœlic spaces. And yet such an arrangement of the tentacles is not known to occur among the corals. It would seem either that the corals are derived from actinine forms with regularly cyclical tentacles, and that the similarities which the actinians mentioned above present to them are due to similar conditions of life, the actinians molding themselves over foreign bodies very much as a coral ployp is molded over its corallum, or that we may yet discover stichodactyline corals. So far as our present information goes we are justified in assuming only an actinine origin for the corals, but if, as suggested, the similarities of Actinotryx and Ricordea to the coral be due to similar life conditions, it would be easy to understand how the formation of a corallum would lead to very general uniformity of structure in forms of different ancestry, and would permit a supposition that the coralligenous forms might have arisen independently from several actinine groups.

A decision on these points must be left for future investigation, which, it is hoped, will be abundantly stimulated by Dr. Duerden's most painstaking and important work.

J. P. McM.

SOCIETIES AND ACADEMIES. NEW YORK ACADEMY OF SCIENCE. SECTION OF ANTHROPOLOGY AND PSYCHOLOGY.

THE regular meeting of the section was held on April 27 in conjunction with the New York branch of the American Psycho

logical Association, Professor Thorndike pre

siding.

Professor E. L. Thorndike reported the results of extended measurements of mental traits in the two sexes. In general the females were less variable. In the case of children 9 to 12 the ratio of female to male variability was .92; in the case of children 13 and 14 it was 1.02; in the case of children 15 it was .97; in high school pupils .95; in college students .85. In the abilities measured the greatest difference found was the female superiority in the tests of impressibility, such as the rate and accuracy of perception, verbal memory and spelling. these only about one third of the boys reach the median mark for girls.

In

Mr. Wm. Harper Davis read 'A Preliminary Report of Tests of Scientific Men,' dealing with some twenty physical and mental measurements made upon one hundred professional men of science, under the auspices of the Committee on Anthropology of the American Association for the Advancement of Science. No significant correlations were found between any of the tests and the several departments of scientific activity, although the cases were too few to warrant an expectation of decided results. (The superiority of psychologists in 'logical memory' was attributed to the accident that the passage used in the tests was psychological in content.) Vivid mental imagery was less common among the older than among the younger men. were detected.

Two cases of color-blindness

Comparison with Columbia College students, upon whom the same measurements have been made, revealed no significant difference between the two groups, except such as would naturally arise from their disparity

in age.

Critical comments were made on some of the tests and on the method of administering them. It is expected that these measurements will be continued under the direction of Professor J. McK. Cattell, who is engaged upon a comparative study of scientific men.

Mr. S. C. Parker presented a paper upon

'Correlation of School Abilities.' Several investigations in Teachers College have had for their subject 'The Correlation of School Marks.' The method and results of these researches are set forth in Vol. XI., No. 2, of the 'Columbia University Contributions to Philosophy, Psychology and Education.' This paper reports the results of some new calculations based on the marks of 245 boys in a New York City high school.

It must be borne in mind that we do not know exactly what school marks represent; they may represent real ability in the school subjects or merely the ability to get marks.

In performing the statistical work, it is important to transmute each teacher's marks separately. This point is mentioned because the neglect of it by one investigator lays his results open to question.

cent.

The

There is not any very great variation in the correlations between marks in academic subjects, such as the languages, sciences and mathematics. The Pearson coefficients run between 40 per cent. and 60 per cent. correlations of drawing with academic subjects are low-lying as a rule between 0 and 25 per From a psychological standpoint, the academic correlations are high. But it must be borne in mind that many constant errors enter in which would make the correlations much higher than the essential relationships would be. From an educational standpoint the correlations are low. They show the futility of the belief in general brightness for all things, and are one of the best arguments for the elective system.

Professor MacDougall read a paper on 'The Specialization of the Hand in Relation to Mental Development.'

JAMES E. LOUGH, Secretary.

ENTOMOLOGICAL SOCIETY OF WASHINGTON.

THE 177th regular meeting was held on April 2, 1903, nineteen members and one visitor present.

Mr. Banks reported that eleven members attended the field excursions to Bladensburg, Md., on March 26. A most enjoyable day was spent and some good specimens secured.