signed by the best authorities on surface tension, are of great interest, but, as they strongly emphasize the shortcomings of the instrument itself, now on the market, they are likely to throw discredit on the method which we maintain to be the best for the study of colloids. Therefore, the writer thought it necessary to add a few words to this discussion. It has never been his intention to claim that the dimensions which he chose for the stock platinum ring were the ideal ones, for all kinds of work. No stock instrument can claim so much, not even stalagmometers and glass tips. For standard work of the highest accuracy, the glass tips have to be carefully calibrated and ground by the experimenter, and are not on the market. The same applies to capillary tubes. For very small values of surface tension, it is advisable to use tips of a different size than those used for water and certain aqueous solutions. This is true of practically every physical apparatus. There is no doubt that a knife-edge ring, such as is used by Dr. Klopsteg and myself in certain careful measurements, is better than the ordinary stock platinum ring with a circumference length of 4 cm. But the tensiometer was made principally to determine very rapidly the surface tension of a small quantity of liquid with accuracy and was particularly intended for the study of the time effect on aqueous colloidal solutions; now, the values obtained for pure water are in excellent accord with those accepted as standards, from which they differ by less than 0.1 dyne. The agreement is better than that which is to be found in the data published by different authors using drop weight methods. This was considered as satisfactory. Dr. Johlin (SCIENCE, 1926, lxiv, 93), acknowledges the fact and explains it by stating that the "approximately correct (?) values found with the ring supplied with the instrument are the result of the cancellation of equal and opposite errors." This is indeed a great compliment to the instrument, in fact the greatest that can be made to any instrument. Further, he states that the value obtained for benzene is too high. Probably he considers the data obtained with the capillary ascension method as the absolute standards. But this method is known to give lower values than the others, and has been seriously criticized by a number of excellent authorities, A. Ferguson among others. In the tables, the surface tension of ethyl alcohol is given as 22 dynes at 20° C. (Ramsay and Shields, capillary ascension), but Grunmach found 26.1 dynes at 17.7° C. (capillary waves), and Freundlich ("Capillary Chemistry," 3d ed., p. 43 of the English translation) quotes 21.6 dynes at 25° C. These values do not agree. When a liquid is in contact with its vapor, the readings are different from those obtained when it is in contact with air. As long as no absolute

theoretical values of the surface tension of pure liquids are available, it is impossible to condemn a method because, under certain conditions, in the case of certain liquids, it does not agree with another.

In addition, I have lately read with great satisfaction a letter by Professor Harkins in Nature, in which he states that he and his collaborators have worked out a correction formula for the ring method reducing the errors to one per cent., in all cases, and that they hope to reduce them eventually to one tenth of one per cent. Such a statement issued by one of the greatest authorities on surface tension ought to settle the question definitely.

Dr. Johlin, in his paper in SCIENCE, evidently aiming to correct the writer, says that "two hours can not be assumed as sufficient for reaching the state of even approximate equilibrium. Frequently the change following an initial period of two hours is several times as great as it was in this initial period." I feel sure that Dr. Johlin will give me credit for not having overlooked such a possibility and that he has understood, as I have, that the time necessary to reach an equilibrium is function of the concentration, of the mobility of the molecules or particles in solution and of the distance they have to travel to reach an adsorbing surface. The latter condition may be Surface expressed by the value of the ratio the imVolume' portance of which has been emphasized in my book. A stable value is attained in less than two hours when 2 cc of a sodium oleate solution at concentrations between 1/25,000 and 1/1,000,000 are contained in watch-glasses; 100 cc of the same solutions will require at least sixty-four hours to reach their equilibrium when placed in a petri dish 10 cm in diameter (see "Surface Equilibrium of Biological and Organic Colloids," p. 174).

P. LECOMTE DU NOUY ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH

THE BASIS OF REFLEX COORDINATION IN SCIENCE, Vol. LXIV, No. 1650, A. Forbes has raised some objections against my theory of specific accord between the excitations sent off by the central nervous system and the motor end-organs. My theory is based on two main points:

3

(1) On the phenomenon discovered by me1,2, and since confirmed by G. Hertwig, Detwiler and W. 1 P. Weiss, Arch. f. mikrosk., Anat. u. Entwicklg. mech., Bd. 102 (635)-1924.

2 P. Weiss, Jour. Comp. Neur., Vol. 40 ( )-1926. 3 G. Hertwig, Sitzungsber. d. natur-forschd. Gesellsch. Rostock, Vol. 1, 192.

4 S. R. Detwiler, Jour. Comp. Neur., Vol. 38 (461)— 1925.

Brandt, that in a supernumerary transplanted limb, when innervated from the limb level of the spinal cord, every muscle enters into action, always at the same time and with the same degree of intensity as does the homologous muscle in the normal limb close to it. It is not quite correct to state, in respect to this phenomenon, as Forbes does, that "the nature of the reflex coordination involved is best illustrated by the fact that in movements of progression all flexor muscles contract together, while the extensors relax, and vice versa," for it is easy to evoke experimentally others than progression reflexes, where not all muscles which are synergic in progression work together; in this case, as well as in the supernumerary limb, not all flexors or extensors are found to exhibit contraction at the same time, but only those among them which are homologous to the normal muscles at work.

(2) On the fact that in innervating the transplanted limb the outgrowing nerve fibers during their course are dividing each in several branches, their subsequent distribution being entirely a matter of chance, as there is not any specificity involved in directing the single nerve fibers. So at least the great majority, if not all, of the motor ganglion cells innervating the supernumerary limb have their several peripheral branches ending on muscles of different kinds.

Forbes admits that if this be really the case my statements in respect to "some power to select a special component in excitation" in the muscle would be correct. But he continues: "Weiss furnishes neither proof nor evidence for his assertion that a single motor neurone may innervate antagonistic muscles. . . . The individual spinal root, containing many axons, may so branch as to supply both the normal and the supernumerary limb, but the individual axon may (and probably does) remain unbranched till it approaches the muscle and there distributes itself only to adjacent fibers."

...

In reality, I did furnish such proofs for my assertion. Every one can see by an exhaustive study of my paper of 1924 that Forbes's conception just mentioned is by no means in accord with the facts or with my statements about these facts. In reconstructing in three animals with supernumerary transplanted limbs the nerve paths, I found and described that the individual axon does not remain unbranched till it approaches the muscle. What really happens is, on the contrary, that the nerve fibers cut off by implanting the limb branch immediately after beginning their outgrowth and are widely distributed long before entering the nerve paths of the limb to be innervated by them. The fiber branches, in running

5 W. Brandt, Arch. f. mikr., An. u. Entw. mech., Vol. 106 (193)-1925.

through the pathless scar, do not at all remain together and when reaching the proximal end of the transplanted limb are so confused that, save in exceptional cases, the order of the fibers in entering the different nerve channels of the transplant is quite other than it was in leaving the central nerve stump. So it is clearly seen that it is quite incorrect to believe the fibers to augment only when they have reached the muscle, as Forbes does. The augmentation takes place long before.

In overlooking this point, it may be easy to give an interpretation of the observed phenomenon on the basis of the classical nervous physiology and there would not be any need to accept my theory. However, recognizing the haphazard disorder of the outgrowing and dividing fibers, as proved by my microscopical examinations, and as recorded in my paper of 1924, we are obliged to accept a resonance-like mechanism involved in the nervous action on the muscle system.

The statement of the older theory that coordination of muscular action is determined within the central nervous system remains untouched by my theory. Only one point must be changed; whereas, after the former theory, the central coordinations were believed to depend on a geometrical distribution of excitation on the paths connected with the muscles to be brought to work at the given moment, it consists, in the new theory, of a dynamical selection of specific excitation forms adapted to the different muscles (the selection may perhaps consist in the excitation of centers which produce discharges just of these forms).

A resonance theory in such a general form as I proposed is the only explanation I can think of which in all respects is in concordance with the facts observed. To bring it in accord with the opinions of nerve physiology generally held will be a matter of future investigation. There are, it is true, many discordances; especially, as is pointed out by Forbes, there is a striking incompatibility between a resonance theory and an all-or-none-principle. But, is the all-or-none principle one of normal reflex action? It may be, as is Forbes's opinion, that there is no definite proof against the assumption that this principle would hold good not only for the inadequate stimulation of the nerve itself, but also for the adequate central innervation. I may point out, however, that, on the other hand, there is no convincing evidence or proof to confirm this assumption.

For all further information I may refer to an extensive publication of my theory which will appear in a few months. PAUL WEISS

BIOLOGISCHE VERSUCHSANSTALT DER
ACADEMIE DER WISSENSCHAFTEN,
VIENNA

"SINGING" EARTHWORMS

AN article in the Literary Digest of October 9, 1926, has been sent to me by Professor Jesse E. Hyde, of Western Reserve University, because he remembered my mentioning the observation of sound-producing earthworms. The article reports, under the heading "When the Earthworms sing Together" the observation of Professor Mangold, of Freiburg, Germany, that "the earthworms possess voices and that they actually are in the habit of uttering slight sounds, and that they do this not singly but in series marked by definite and varying rhythm."

Seeing that the fact that earthworms make noises had not been known before, as I had assumed, I wish to record the observation that also American earthworms produce sound.

It was first pointed out to me by Mrs. Ruedemann about a decade ago, on a sultry May evening, that the earthworms in our garden back of the house could be distinctly heard. Being incredulous at first, I sat quietly on a chair until I also heard an exceedingly fine rasping noise all around me. It was a chorus of almost unbelievably small voices in the dark. To find out whether the little musicians were really earthworms, I got a flashlight and when the voices, after the quiet resulting from the disturbance of walking over the ground, were again in full chorus, turned the light upon a point close to me, from which I was sure a rasping sound arose. The light revealed a large earthworm, partly stretched out of its burrow. I spotted several more afterwards. We two have since heard the singing every year, always on warm spring evenings about and after dusk. Mrs. Ruedemann also heard it last spring about 4 o'clock in the afternoon on a warm May day after a rain, and then she could see the "singing" worms all partly stretched out of their burrows.

From the rasping character of the sound and the position of the worms I inferred that the noise was made by the drawing of the setae over some hard object at the edge of the burrow, and the time of the year suggested that the concert is connected with the mating season of the worms. Professor Mangold, on the other hand, concludes that the sound is made through the mouth and is more of the character of clicks, which however may "sometimes become so rapid as to form a buzzing noise." These noises were made only in the burrows in his aquarium.

A member of the museum staff, Mr. Jacob Van Deloo, tells me that he heard the sound frequently, when a boy.

Not being aware that this "musical talent" of the earthworms was unknown to naturalists, I failed to catch some of the musicians for identification. Dr.

ENGLISH VERSUS METRIC SYSTEM

THERE has been considerable agitation for replacing our English system of weights and measures with the metric system.

One of the most striking examples of the unscientific way in which the English system of weights and measures has grown up and the confusion that it introduces is given in the following sentence taken from a paper on "The Effect of flooding with Sea Water on the Fertility of the Soil," by H. J. Page and W. Williams, of the Rothamsted Experimental Station, Harpenden, England, and published in the Journal of Agricultural Science, Vol. XVI, Pt. 4, pp. 551-573 (1926).

The land is typical strong wheat and bean land which can ordinarily be expected to give a yield of four to five quarters of wheat per acre.

I am supposed to be familiar with the English language, and yet the quantity, "a quarter of wheat," was a new term to me. I accordingly looked it up in Funk and Wagnall's New Standard Dictionary (unabridged) and found that the following possibilities present themselves.

(1) The fourth of a hundred weight (this would mean 25 pounds).

(2) By the old reckoning, the fourth of a hundred weight, where the hundred weight is 112 pounds, namely, 28 pounds.

(3) Eight bushels, with the parenthesis following it (in some localities, 8 3/4 or 9 or 12 or 16 bushels, etc.).

(4) A fourth of a ton. (Query: Is a short ton, 2,000 pounds, or a long ton, 2,240 pounds, meant?)

Apparently the dictionary had not helped me very much in deciding what the authors meant by "a yield of four to five quarters of wheat." It would have been just as intelligible to me if they had stated "two or three cart loads" and had neglected to state the size of the cart. Accordingly, I asked some of our graduate students from Canada what was meant and they said that we would have to find out what a "quarter" meant at the particular grain market where the wheat was sold, in order to decide what the authors meant in this scientific paper.

If the English system of weights and measures can

SCIENTIFIC BOOKS

Cloud Studies. BY ARTHUR W. CLAYDEN, M.A. Second Edition. E. P. Dutton and Co., N. Y. STUDENTS of nature should be pleased that another edition of Clayden's "Cloud Studies" is now available. Mr. Clayden has given a great deal of time to developing the art of cloud photography and the result is a series of beautiful cloud photographs which should introduce any reader to a knowledge of the different cloud forms. These cloud forms are called by the names adopted at the International Meteorological Conference at Munich in 1891, a modification and extension of the cloud nomenclature introduced by Howard in 1803.

Mr. Clayden has been an enthusiastic observer as well as a photographer of clouds and in his introduction he tells how to observe clouds easily by means of a blackened mirror. Such a mirror diminishes the glare and brings out in a wonderful manner the detailed structure of the finest cirrus and enables one to observe right up to the edge of the sun. It enables one to view the clouds looking downward instead of in the unnatural position of stretching the neck and the face upward. In the more comfortable position of gazing downward into the mirror long-continued observations may be made and one form of cloud can be watched changing into another.

Beginning with the highest cloud forms Mr. Clayden takes up in succession the different forms of clouds beginning with the highest. In chapter III he pictures, describes and names no less than nine forms of cirrus, quite distinct from each other; but some of these are transition forms and border closely on cirro-cumulus and cirro-stratus. Chapter IV is devoted to cirro-stratus and cirro-cumulus, and numerous examples of each form are given. It is difficult to photograph the widely extended sheets of cirrostratus so that most of the photographs partake of the cirro-cumulus type. Chapter V takes up the "Alto" clouds, alto-stratus and alto-cumulus. It is

almost impossible to photograph the widespread, almost uniform, dark sheets of alto-stratus so that most of the examples given are of alto-cumulus. Chapter VI is devoted to the lower clouds, stratus, strato-cumulus and nimbus. The different types are illustrated by photographs, but these are much less satisfactory than those of the upper clouds. Owing to the absence of color in the photographs it is difficult for an inexperienced person to tell whether dark patches are clouds or sky. This difficulty can be overcome only when it is possible to photograph in colors. In chapter VII he takes up the cumulus which is perhaps, the best known cloud and easy to photograph. In chapter VIII is described the cumulonimbus or shower cloud. Some photographs show the anvil-shaped top of the cloud and others the massive cumulus-like structure of the cloud. Here again we miss the absence of color to distinguish cloud from sky. In chapter IX he discusses clouds which form in wave-like lines and ripples and finally in chapter X has an excellent description of methods of photographing clouds and of determining cloud heights by photography.

The strength of the treatise lies in its photographs and descriptions of clouds. Mr. Clayden has clearly been a student of nature and not of books. His discussion of how clouds are formed and the physical processes involved is very inadequate. It is true that he rightly attributes cloud formation to adiabatic cooling of moist air by expansion and refers to the work of Aitken and Wilson as to the necessity of nuclei for condensation of moisture in droplets. He makes no mention, however, of the usual formation of cloud sheets in inclined strata and ignores the work of Bjerknes and other writers who show that the chief cause of cloud formation and of the ascent of air in inclined strata is the contrast of adjacent bodies of air at different temperatures and the overrunning of colder air from the direction of the pole by warmer air from the south and east.

H. H. CLAYTON

Die Vögel Mitteleuropas. Herausgegeben von der Staatl. Stelle für Naturdenkmalpflege in Preussen. By DR. OSKAR and FRAU MAGDALENA HEIN ROTH. Hugo Bermühler Verlag, Berlin-Lichterfelde. Lieferungen 1-10; 1924-1925; pp. 1-80; 16 colored plates; 42 black plates.

THE first ten parts of this work appeared between July, 1924, and April, 1925. They are devoted to an account of part of the order Passeriformes of central Europe, i.e., Germany, and they include the wren, water ouzel, accentors, thrushes, flycatchers, waxwing, shrikes, swallows, and the beginning of the Old World warblers, in all thirty-one species.

There is given a brief general statement regarding the families to which these belong and the more or less convenient groups into which the authors divide some of the families. There is also of each species an account, which, while it does not go into technical details, gives habits, measurements, and geographic distribution, though usually not a complete description of adult plumages. While no attempt is made to provide a full life history, the material given throws much interesting light on the birds treated, and in some cases is considerably extended. The text is made up in large part of original life history observations of birds in the field and in captivity, and for this reason the accounts of the various species differ much in length, according to the opportunities of the authors.

Much exceedingly interesting information is given on the song, notes, nesting habits, the young, and the general behavior of the different birds. In the exactness of the data presented, the text gives excellent evidence of the care with which the observations have been made. Of particular value are the notes on the development of nestling birds, with specific data on the age of each stage of plumage, information which every one who has had occasion to search for realizes is difficult to obtain, and furthermore, all too rare in books on birds. Dr. Heinroth has for a great many years been able to rear large numbers of birds in captivity, and this has given him an unexampled opportunity to determine many facts which would easily escape the chance observer in the field, but which are of great importance, nevertheless. Some of the longer accounts of the life history of the species treated are especially good and amount almost to a monograph of the behavior of the species. Worthy of particular mention are those of the nightingales (Luscinia megarhyncha and Luscinia luscinia), the blackbird (Turdus merula), the spotted flycatcher (Muscicapa striata), the robin (Erithacus rubecula) and the swallow (Hirundo rustica). The book is not a technical treatise, but is designed to present an interesting side of the life history of the birds of the author's region.

One of the most important features of "Die Vögel Mitteleuropas" is the wealth of illustration. The numerous plates are, however, issued apparently as material was obtained for their completion, or as convenience dictated, since but in few instances do they illustrate the species described in the parts in which they appear. Both the colored and the uncolored plates are made up with the idea of showing the development of the species figured, from the egg through the nestling and juvenal plumages up to that of the adult. The value of these plates is greatly enhanced by the circumstance that they represent the

various stages of plumage at precisely known ages of the birds, in this admirably supplementing the descriptions given in the text.

If the standard already set in the parts now discussed should be maintained, as we have every reason to believe it will be in the parts yet to be published, we are probably not saying too much when we predict that this work will prove to be one of the most important contributions to the life history and behavior of European birds that has appeared in many years. HARRY C. OBERHOLSER

SPECIAL ARTICLES

THE FIBRILLAR STRUCTURE OF THE DENTAL ENAMEL MATRIX OF THE GUINEA PIG

RECENTLY We1 have outlined certain morphological findings with regard to the organic matrix of guinea pig dental enamel. Preparations were made from material carefully decalcified through celloidin. By such a method, sections of this structure may be demonstrated.

Carter2 has claimed that in the formation of enamel, globular material is laid down irregularly and "that there is no sign of any merging of the cells into the secretion such as one would find did the ameloblasts themselves become transformed into a stroma which became incorporated into the enamel." Our sections show that protoplasmic processes may extend from the ameloblastic layer into the enamel structure. These taper away to a point within a distance of from ten to fifteen microns. Thus, there is in the guinea pig a definite articulation between these two elements, as shown by Van Giesen stain.

Moreover, when the matrix is drawn away from the ameloblasts by the pull of the microtome knife, the organic matrix may split or tear. This rupture, however, is in a plane parallel to the length of the enamel rods. Fibril-like structures thus are formed. This result is produced likewise when the direction of tension does not coincide with the length of the enamel rods but lies at an angle with them. We have photomicrographic evidence of these observations.

These results tend to show that a protoplasmic fibrillar structure may connect the ameloblastic layer with enamel matrix in the guinea pig and that the matrix itself appears to be of fibrillar structure. Thus, we do not coincide with the belief that the formation of dental enamel is an irregular deposition of precipitated material. Rather, precipitation of calcareous 1 Beckwith, T. D., and Williams, A., Proc. Soc. Exp. Biol. and Med., 1926, 24, 76.

2 Carter, J. T., Quart. Journ. Micros. Soc., 1918-19, 63, N.S., 387.