inlets are usually broader at the top, and narrower below, with the result that the waters enter and leave the lagoon most freely when the tide is high. As a consequence, the discrepancy between high tide levels in the two water bodies is not so great as that between the low. tide levels; and the mean level of the lagoon is therefore higher than the mean level of the Changes in the number or breadth of inlets much cause changes in the mean level of the lagoon; and such changes are of common occurrence.

ocean.

The shallowing or deepening of inlets; the growth of bars across the mouths of bays formerly free from such shore features, or the destruction of bars by storm waves; the narrowing of inlets by sand-spit growth, their widening by wave or current action, or the breaching of bars by new inlets formed by storm waves or by the outburst of impounded land waters; any and all of these must be potent causes of local changes in mean sea-level in harbors, bays or lagoons where are found conditions approximating those described above. Nor do the conditions described exhaust the list of those which may give rise to local differences in mean sea-level. They are merely examples intended to illustrate the fundamental principle that local changes in the form of the shore may, under appropriate conditions, produce local changes in mean sea-level. Such changes in mean sea-level may be gradual or sudden, depending on the nature of the shore changes responsible for them; and they may amount to fractions of an inch or to a number of inches, depending on the form and size of inlets and bays and on the range of the tides. Where gradual and imperceptible, yet of significant amount, they are apt wrongly to be attributed to progressive, slow coastal subsidence or coastal elevation.

COLUMBIA UNIVERSITY

DOUGLAS JOHNSON

QUANTITATIVE DETERMINATION OF
ROCK COLOR

THE need for standardization of rock colors has been realized by many petrographers. Sedimentationists, especially, have desired a color standard. In meetings of the Sedimentation Committee of the National Research Council, the possibility of basing important deductions as to alteration and environment of sediments upon slight color differences has been suggested. The difficulty in investigating these suggestions has been the lack of means of detecting the requisite small color variations. The best standard of colors now in use is the Ridgway chart. In fact, its use in sedimentation has become so desirable that the Sedimentation Committee has taken steps toward the preparation of a more simple and cheaper chart, based on that of Ridgway, but especially

adapted to field and laboratory descriptions of sediments.

Because the writer does not know of any application of quantitative color measurements to geologic investigation, he believes that the attention of mineralogists, petrographers, sedimentationists and others interested in color work should be called to the fact that instruments are available by means of which colors can be analyzed and synthesized quantitatively. No details of the construction and manipulation of these colorimeters need be given here for this information can be supplied by the dealers selling these instruments. Their wide range of application is indicated by their successful use in industrial plant control and research in a variety of industries including dye, paint, varnish, ink, and soap manufacturing, sugar refining and other industries. Since these instruments have proved their usefulness in practice, it is believed that they will be found to be useful also in the field of pure science wherever color is involved. Their value in petrographic research has been demonstrated by the writer in his study of the relationship between structure and color of the shales of the Cromwell Oil Field of Oklahoma.

The most simple and obvious method of determining rock color by direct comparison of rock fragments with a standard color chart is at best only qualitative. Comparisons of streaks produced in the usual way by drawing fragments of the material over an unglazed porcelain plate enable smaller color differences to be detected than is possible with the use of chips but this method is also unsatisfactory. Streaks vary with slight differences in hardness and texture of the rock as well as with differences in composition. The texture, hardness and whiteness of the streak plate and the pressure applied in obtaining the streak are also significant variables. Moreover, such a streak can not be representative of the sample as a whole because it involves too small a quantity of material and it has the added disadvantage of being neither reproducible nor easily preserved for future reference.

Many of these difficulties can be overcome, as was done in the study of the Cromwell shales, by selecting an average sample, grinding the rock and sieving it. A portion of the powder passing the one sixteenth millimeter screen can then be pressed into a cardboard frame, previously mounted on a datum card, and covered with an ordinary thin cover glass. Black binding tape such as is used in preparing lantern slides serves to hold the glass to the rest of the mount. Such a record is permanent and can be filed. It overcomes the objections of the above-mentioned methods, but it too has some disadvantages. First, the mounting of the powders is time-consuming and,

secondly, after all this work has been done, the color differences detectable by the eye are not always sufficiently small to be of value in the investigation. The fundamental need, a method of greater sensitivity, remains.

Such was the case in the Cromwell problem when a timely advertisement in SCIENCE called the writer's attention to the color photometer. In order to test the applicability of the instrument to the study of the shales in question, the writer submitted samples of shale powders passing the one sixteenth millimeter screen to the dealer for trial tests. The results agreed so well with certain chemical determinations that the writer believes that he is warranted in suggesting the use of the color photometer in other investigations. When the investigation of the Cromwell shales is completed, it is hoped that the application of quantitative color data to petrographic research will be demonstrated conclusively. As stated in the beginning, the purpose of this brief paper is merely to make better known a color-determining device applicable to liquids, powders and massive solids, both heterogeneous and homogeneous, capable of giving quantitative data which can be presented graphically. Such an instrument may prove of great value in other geologic problems such as those dealing with changing environments under which sedimentary beds have been deposited, color changes produced in rocks during metamorphism, and in other types of investigation. The color of mineral streaks can now be placed on a quantitative basis. Doubtless, other applications will suggest themselves to the reader. OLIVER R. GRAWE

MACKAY SCHOOL OF MINES, RENO, NEVADA

...

A NEW FUNDAMENTALIST STRONGHOLD "THE Des Moines University, Des Moines, Iowa, is now the property of the Baptist Bible Union of North America . . . A President has not been elected, but in the meantime the Board of Trustees announce that no one will be retained on the faculty who is not a Christian in the sense of having been born again Some professors will teach no longer in the university because their views are decidedly modernistic... No professor will be retained who believes in evolution, or who does not accept the Bible as the infallible word of God. . . The highest educational standards will be maintained... Des Moines University will teach the supernaturalism of Christianity as opposed to the naturalism of modernism which is prevalent to-day."

The above, taken from a publication of the Baptist Bible Union, is published because the situation should

be thoroughly understood by scientific men. Twenty of the faculty, including two deans, have resigned. The writer, who a year ago accepted a two-year contract as professor of biology, with the promise of freedom in the teaching of evolution is among those leaving. N. M. GRIER

HATHAWAY PARK,

LEBANON, PA.

QUOTATIONS

STEEL TURNS TO RESEARCH

SCIENCE is to work for the United States Steel Corporation. To be sure, the greatest organization of its kind in the world has long had its laboratories, but it has been their main function to make more or less routine analyses and to control the processes whereby ore is converted into hundreds of products ranging from wire to girders. No startling discovery in the chemistry of iron and steel stands to their credit. The

corporation has made its greatest technical strides in engineering—in lowering production costs, in introducing new machinery, in increasing tonnage. Convinced, no doubt, by the example of other large industrial organizations and above all by Sir Robert Hadfield, of Sheffield, and the great German ironmongers, the United States Steel Corporation has decided to create a department of research and technology under the direction of Dr. John Johnston, of Yale, a scientist ably qualified by technical education and experience to explore a field in which scientific and in

dustrial honors are to be won.

Judge Gary's announcement of what his board of directors must have regarded as a daring innovation is phrased with characteristic but guarded optimism. The finance committee is to keep an eye on the research laboratory. While the corporation "has no money to waste intentionally," Judge Gary comments, "we have money to expend if necessary." Miracles are not to follow the rubbing of the lamp of science by a chemical Aladdin. "We do not expect you can go along at a very rapid rate to begin with, or, perhaps, at any time, but we will have patience, as you must all have patience."

Some research is better than none, particularly if the spirit in which it is conducted is that of the university. How successful the new department of research is destined to be must depend largely on the policy adopted. Such experienced directors of research laboratories as Dr. W. R. Whitney, of the General Electric Company, and Dr. C. Kenneth-Mees, of the Eastman Kodak Company, have argued for an absolutely free hand. Money-making must not infect the laboratory. Paradoxically, the most money is

made by laboratories least concerned with it-by men who have dabbled in the Einstein theory and the mysteries of the Bohr atom and stumbled on principles applicable to industry. If purely commercial standards are to guide the research director he finds it difficult to attract men of the finest scientific type. His net result is merely a heightening of technical efficiency, an improvement in finished products. Grant a laboratory the right to work untrammeled and both science and industry gain. It was the adoption of this large-visioned policy that made the discovery of ductile tungsten possible-a discovery that unexpectedly gave us electric lamps of an economy and brilliancy undreamed of twenty years ago, radio tubes that have made broadcasting and television twentieth century triumphs, and deeply penetrating X-ray tubes that have been a boon to the sick.

The richest assets of some of our largest corporations are not their physical properties but the discoveries made in laboratories where research has been conducted for its own sake. Perhaps because these assets can not be even approximately appraised, at least one corporation carries its priceless patents on its books at the valuation of one dollar.-The New York Times.

SCIENTIFIC BOOKS

The Ferns (Filicales). Vol. II. The Eusporangiatae and other relatively Primitive Ferns. F. 0. BOWER, SC.D., LL.D., F.R.S., pp. 344, many figures. Cambridge, the University Press. 1926. FOR more than forty years Professor Bower has been recognized as a leader in the study of the Pteridophytes; and this work, the second volume of a comprehensive treatise on the ferns, of which the first appeared in 1923, is especially welcome to those, who in these days when morphology is rather discredited still feel that the subject not only is far from exhausted, but will again be revived when some of the current botanical fashions are out-moded.

The present volume treats in detail the Eusporangiatae and the more primitive families of the Leptosporangiates, and is a contribution of the first importance. It records the latest conclusions of the author as to the structure and classification of the ferns.

gium-development, and on this basis he arranges the families in three categories, viz.: Simplices, in which all the sporangia of a sorus are formed simultaneously; Gradatae, in which they are of different ages, formed in basipetal succession; and Mixtae, in which sporangia of different ages are mingled in the same sorus. The Simplices are the most primitive, the Mixtae the most specialized.

There are two types of sorus, marginal and superficial, i.e., borne on the lower surface of the leaf. The marginal sporangia are believed to be the older type, although the superficial sori are characteristic of the Marattiacae as well as of some other paleozoic ferns. The present volume deals with the Simplices and Gradatae, of which fourteen families are recognized.

Before considering the living ferns, a chapter is devoted to a group of fossils, Coenopteridaceae, which have no existing representatives. There are three families of these: Botryopterideae, Zygopterideae, and Anachoropterideae. They are all confined to the Palaeozoic, occurring from the Upper Devonian to the Permian.

The author concludes that the Coenopteridaceae include an assemblage of more or less synthetic types which may probably be assigned to the Filicales, but which do not show any close relationships with existing ferns.

Of the living Filicales, it is pretty generally admitted that the two Eusporangiate families, Ophioglossaceae and Marattiaceae, are the most primitive.

In his earlier writings Professor Bower separated the Ophioglossaceae from the Filicales, but in the present work he has restored them to a place among the ferns, where there is no doubt they belong. It is true that their exact relationship with the other ferns is not easy to determine.

While almost nothing is known of the geological history of the Ophioglossaceae, there is very strong evidence that they are the most primitive, and presumably the oldest, of the living ferns. There seem to be sufficient resemblances, to the fossil Coenopterideae to warrant the assumption of a remote relationship with that order.

midrib and lateral veins like those of the typical ferns. Now the transition from the cyclopteroid venation of Eu-botrychium to the simple reticulate venation of the cotyledon of Ophioglossum Moluccanum, for example, is not a very great one. A similar transition from the open venation to the reticulate is shown by Professor Bower in Marsilea (p. 179, Fig. 461). In short, the contrast between the venation in Ophioglossum and Botrychium is not so marked as Professor Bower believes.

The statement (p. 57), "In the ontogeny of the Pteridophytes a coherent body of tissue called the stele, partly made up of elements having a truly cauline origin, exists from the first, and it serves to connect up adjacent leaf-traces," is certainly open to question. A most careful study of the ontogeny of Ophioglossum, especially O. Moluccanum, has shown as conclusively as possible that the whole of the vascular skeleton of the axis is of foliar origin and that there is no truly cauline stelar tissue. This is true also for Botrychium and probably for Helminthostachys, as well as for the early stages, at least, of the Marattiaceae.

It may be said that Professor Bower seems to be aware of the difficulty in harmonizing the stelar theory with the conditions that exist in Ophioglossum.

Professor Bower's studies on the development of the sporangium in the Ophioglossaceae are quite the most complete that have been made, and are amply treated in the present volume. One may venture to differ from his conclusions in one particular, viz., the nature of the sporangial spike. There is good evidence that this is not an appendage of the leaf, but a structure coordinate with the whole sterile segment. Both in Ophioglossum Moluccanum and Botrychium Lunaria there is a dichotomy of the very young leaf primordium, the branches forming respectively the fertile and sterile segments.

A sufficiently complete account of the gametophyte is given, but the embryo and young sporophyte, especially in Ophioglossum, are not treated as fully as might have been wished. Why the young sporophyte in O. Moluccanum, with its functional cotyledon, should be considered less primitive than that of the other species in which the early leaves are rudimentary, is hard to understand; nor will the conclusion that Ophioglossum is less primitive in structure than the other genera be accepted without question. Space will not permit a fuller discussion of these points.

The very distinct order Marattiaceae is of particular importance in the phylogeny of the ferns, since unlike the Ophioglossaceae, to which they are undoubtedly related, there are abundant fossils obviously allied to living forms.

In the later Palaeozoic, fern-like fronds with sori similar to those of existing Marattiaceae are found,

and in the older Mesozoic rocks occur fossils much like the living genera.

The statement (p. 102) that the very young sporophyte of Danaea is "protostelic," is incorrect, as there are several distinct xylems belonging, respectively, to the leaf traces which have united to form the solid stele.

The relationships of the Marattiaceae to the other ferns are difficult to determine. They seem to be relics of a Palaeozoic and Mesozoic stock which have come down to the present with little change and have not given rise, directly at least, to any of the existing Leptosporangiates.

In the enumeration of the number of living Marattiaceae (pp. 124-125), there is an obvious typographical error. Christensenia (Kaulfussia) has only two species, not 26, as indicated in the table.

To some extent intermediate between the true Eusporangiatae and the Leptosporangiatae is the small family Osmundaceae with two genera, Osmunda and Todea, and 17 species. Like the Marattiaceae, the living species are but remnants of a once much more extensive order. The earliest fossils of Osmundaceae are in the Permian, where perfectly preserved stems closely resembling the structure of living species are found.

The intermediate character of the Osmundaceae is shown in the gametophyte, embryo and sporangia, as well as in the anatomy of the adult sporophyte. This is excellently summarized on page 148.

The three remaining families of Simplices, Schizaeaceae, Gleicheniaceae and Matoniaceae, like the Osmundaceae, are undoubtedly relics of once much more predominant types. Of these the Schizaeaceae lead up to the series of Leptosporangiates with marginal sori, while the Gleicheniaceae are the most primitive of the series with superficial sporangia.

The Gleicheniaceae, and the related Matoniaceae, are very uniform in their structure; but the Schizaeaceae differ greatly among themselves, and their relations to the other ferns, both living and fossil, are by no means clear. Possibly going back to the Carboniferous, and certainly to the Jurassic, they show great variety both as to external form and anatomy. Their sporangia, however, are quite uniform in type.

Probably an offshoot of the Schizaeaceae are the heterosporous Marsileaceae, which agree closely with the Schizaeaceae in their anatomy and in the development of the sporangia.

During the Mesozoic, especially the Cretaceous, species of Gleichenia were abundant in the northern regions, extending even to West Greenland. Fossils resembling Gleicheniaceae occur in the coal measures, but there is some doubt as to their real nature.

The first family of the Gradatae, the Hymenophyllaceae, is a very natural one, all of the nearly 500

species being referable to the two closely related genera, Hymenophyllum and Trichomanes. Their geological history is obscure, but the latest conclusion is that the family is not an extremely old one. Their nearest relationship is probably with the Schizaeceae. Formerly included in the Hymenophyllaceae is the monotypic genus Loxsoma from New Zealand; but it has now been separated as the type of a separate family, Loxsomaceae, which also includes three species of a recently described second genus, Loxsomopsis. Professor Bower believes that the Loxsomaceae are related to the Dicksoniaceae.

The most radical change in classification is the separation of the Cyatheaceae, to which most of the tree-ferns belong, into three families, viz., Dicksoniaceae, Protocyatheaceae and Cyatheaceae, the latter including only the three genera, Cyathea, Hemitelia and Alsophila. The Dicksoniaceae have marginal sori, and are believed to have no relation to the Cyatheaceae, in which the sori are superficial The family Protocyatheaceae is proposed to include two genera, Lophoria and Metaxya.

Professor Bower notes a remarkable peculiarity of the young sporangia in Metaxya and the Cyatheaceae in which they differ from all other ferns that have been investigated, viz., the apical cell of the young sporangium is two-sided, instead of three-sided. Figure 55 suggests the segmentation in the antheridium of a moss.

The family Plagiogyriaceae is proposed to include the single small genus Plagiogyria. It is to some extent a synthetic type, intermediate between the Gradatae and Mixtae. "It is a relatively primitive type, but not very closely allied downwards to any one of the known primitive Ferns."

The last family discussed in the present volume is the Dipteridaceae, with the single genus Dipteris, as to whose relationship there has been some controversy.

The final chapter is an excellent summary of the conclusions reached from the detailed study of the different families. This chapter includes maps showing the present distribution of several of the most important families, as well as their occurrence in a fossil condition. There is also a diagram showing the relationships of the families discussed in the text. Professor Bower's long.continued and exhaustive investigations in the development of the sporangium have made him the leader in this important subject, and he has treated it admirably in the present volume. It is this perfect mastery of the subject which makes his classification, based mainly upon sporangial characters, so satisfactory. There will probably be little dissent from his conclusions.

One could wish that less space had been devoted to the elaborate details of stem-anatomy, and somewhat

more to the gametophyte and embryo-sporophyte, especially to the question of the origin of the vascular system.

The conclusions reached by recent studies on the origin of the vascular tissues of the Eusporangiatae point to a foliar origin for the bundles of the axis, and these results are hardly given adequate attention by Professor Bower. It is by no means unlikely that further investigations on the vascular bundles of the Leptosporangiates will show that in them also, there is no "stele" in the sense used by the author.

Professor Bower is to be congratulated on the completion of the second volume of this very important undertaking, and the final one will be looked forward to with the keenest interest. To all students of the Pteridophytes these volumes will be indispensable.

DOUGLAS HOUGHTON CAMPBELL

STANFORD UNIVERSITY, CALIFORNIA

SPECIAL ARTICLES

THE INFLUENCE OF X-RAYS ON THE DEVELOPMENT OF DROSOPHILA LARVAE

DURING the past two years we have been engaged in carrying out experiments the results of which we have hoped would give some definite data concerning certain fundamental aspects of radiation effects on biological processes. As a preliminary report we wish to present some of the observations made on the influence of X-rays on a given biological process, namely, the development of Drosophila larvae into pupae. The larvae employed in our experiments have been raised from an original culture of Drosophilae obtained from Dr. J. H. Northrop who had grown these flies under aseptic conditions for many generations; and we have maintained the same conditions.

Our procedure, briefly stated, has been to wash larvae (mean age, 2.5 days) out of a seeding flask on to a piece of aseptic voile, then to transfer them by a method of random sampling to wells in paraffin blocks, or to paraffin permeated pill boxes (in which case the boxes were then set in wells in paraffin blocks). A Kelly-Koett X-ray machine, supplied with 12.5 cm. spheres for spark gap, has been used throughout.

We observed that the larval stage was significantly prolonged, and that the fraction of the total number of irradiated larvae reaching the pupal stage was sensibly the same as for controls, when the conditions of irradiation were as follows: Spark gap 2 cms. distance between spheres; M. A., 8; target distance, 30.5 cms. Three experiments were then performed in each of which three lots of larvae were irradiated