the polarities shown, the direction of the e.m.f. induced in B is "in," as indicated by the cross, the filaments f and f' acting concurrently. This agrees with the general law that when the primary current decreases, the secondary induced e.m.f. is in the same direction as the primary current.

The tube A may be considered as consisting of pairs of filaments, such as f and f'. Since an elementary e.m.f. is induced in B by each pair of filaments, and the action is cumulative, a finite e.m.f. should be induced in B when di/dt in the whole tube has a finite value.

Thus, according to Reasoning I, there should be no e.m.f. induced in B, while according to Reasoning II, there should be an induced e.m.f. of finite value. Before unraveling this seeming paradox, the following propositions should be considered:

(1) Is it legitimate to speak of an e.m.f. induced between the open ends of a long straight conductor? To measure this e.m.f. it would be necessary to introduce leads to a voltmeter, thus forming a closed circuit. If an electrometer be used instead, the circuit would still be closed through electrostatic lines of force within the instrument. Should the leads and the measuring instrument be placed within the tubular conductor A, there should be no indication when the current i is varied. Should the instrument and the leads be placed outside A, a loop would be formed, linking with some of the external flux H, and the induced e.m.f. would depend upon the total flux enclosed by the loop.

(2) Careful writers do not speak of an e.m.f. induced in an open straight secondary conductor, but of the direction of the secondary current. This implies a closed secondary circuit and avoids the vexed question as to the seat and location of this e.m.f. See, for example, J. C. Maxwell, Electricity and Magnetism, Vol. II, p. 178; Foster and Porter, Electricity and Magnetism, p. 394.

(3) In Fig. 3, let K be a straight infinite conductor carrying a current i. Let N be a parallel secondary conductor of finite length, with open ends, at a distance r from K. Let the current i return through a cylindrical shell P of very large radius R.

The lines of force due to i are concentric circles, and the flux 9, comprised between N and P, per unit of axial length, is proportional to i log (R/r). Should i vary at the rate di/dt, the e.m.f. induced in N, per unit length, would be proportional to (di/dt) log (R/r). But R is arbitrary and tends to infinity, so that the e.m.f. induced in N seems to be indefinitely large. Here again, to measure this e.m.f., the circuit of N would have to be completed, for example by means of a parallel wire N', at a distance r'. The flux enclosed in this secondary loop has a finite value,

proportional to i log (r'/r), and the e.m.f. induced in the loop (not in one of the conductors) has a definite value (finite) confirmed by experiment.

(4) If an e.m.f. could be induced in a long straight secondary conductor, as shown in Figures 1 and 3, then by grounding one end and providing the other end with a sharp point, an intense local electrostatic field should be produced. The existence of this field could perhaps be demonstrated by some delicate ionization experiment, Stark effect, etc. On the other hand, grounding one end would give a closed circuit, through displacement currents along lines of force between the sharp point and the ground, so that the experiment may not be conclusive.

Thus, on the whole, it seems as though the foregoing paradox is based on the impossibility of either computing or measuring an e.m.f. induced in an open conductor, without considering a return circuit of some kind, either conducting or through a dielectric. In view of the very fundamental nature of the phenomena and laws involved, it is hoped that other points of view will be contributed to this discussion. VLADIMIR KARAPETOFF

CORNELL UNIVERSITY

RATE OF VIRUS SPREAD IN TOMATO PLANTS

WHEN a plant is inoculated at one point with a virus disease, at what rate does the infective principle diffuse itself to other stems, leaves or shoots? Assuming that the incubation period is constant-that symptoms will appear in a given time after the infective agent has reached any point-the appearance of symptoms in a succession in other portions of the plant distant from the point of inoculation ought to provide a measure of the rate of virus spread from the original inoculation point. This observational method, however, relies on uniformity of growth in all parts of the plant and such uniformity may not exist; it further depends on the detection of symptoms at the same stage in their development, which is by no means a certain procedure.

The more direct method of measuring the progress of virus in a plant system here outlined appears to avoid the disadvantages mentioned and to provide a means, accurate within certain limits, of measuring the rate at which the virus moves from part to part of the plant. The results of the short series of preliminary tests are here recorded largely for the purpose of calling attention to and illustrating the method, since the conclusions that might be drawn from the few cases under observation must necessarily be accepted as only a rough approximation to the truth.

Eight tomato plants in pots were grown in such a manner as to develop several horizontal branches, each

of which was bent and led under the earth in a secondary pot to encourage rooting and thus form a readily detachable second plant. The rooting process was hastened by a partial cut between the original and secondary pots. There was thus produced a "colony" with all its units organically connected but capable of being separated at any time and in any fashion desired. The colonies were grown in a greenhouse under a close cheese-cloth cage. The greatest care was taken throughout to avoid accidental infection through insects, handling, touching of leaves, watering, etc. There is no evidence that any such accidental infection occurred anywhere in the series.

When all secondary plants were well rooted but still attached to the parent plant a single shoot of the parent was inoculated with freshly expressed juice from tomato leaves showing marked mosaic. A glass tube drawn to a capillary point was used for the purpose, pressure being supplied by means of a dropper bulb on the end. Inoculations were made near the growing point.

After inoculation a single secondary plant was removed from each colony at intervals of three, ten, fifteen, nineteen and twenty-four days where the number of daughter plants was sufficient for such a series. These isolated plants were kept under observation to 1 see if mosaic developed.

Twenty-four days after inoculation a record of the various series indicated that in two colonies (B and G) the inoculation had failed. There was no sign of mosaic in the shoot originally inoculated or any of the daughter plants in either colony. In the remaining six all plants removed after nineteen days had marked mosaic symptoms on the young growth; in five of the six the disease had appeared in plants removed after

fifteen days; and in three plants taken away after ten days the disease was also evident. None of the plants removed after three days had developed mosaic twenty-four days after inoculation.

It is evident from the above results that the infective principle was unable to pass from the point of inoculation beyond the place of separation in any case in three days; that in half the cases not more than ten days was required to traverse this distance; that in five out of six cases the virus had passed into the daughter plants in less than fifteen days; and that in only one case was a period of fifteen days insufficient. In this case the two plants removed after nineteen days were both affected by mosaic on the twentyfourth day, so that if one allows for a suitable incubation period it is evident that the point of separation must have been passed near the fifteen-day period.

The distances to be traversed by the virus in these colonies varied from eight to eighteen inches. We may see from the above records that these distances were traveled by the virus in periods which might be something less than ten days or slightly more than fifteen days. We have no right to assume that a uniform advance was made during this period, but for purposes of expressing the rate of progress of the virus in concrete fashion it may be permissible to adopt the average rate in common usage for such purposes. On this basis the transfer of mosaic virus appears to take place through the shoots of the tomato plant at a rate somewhere in the neighborhood of one to two inches per day or one to two millimeters per hour.

PENNSYLVANIA BUREAU OF PLANT INDUSTRY, HARRISBURG

W. A. McCUBBIN, F. F. SMITH

FEEDING PLANTS MANGANESE THROUGH THE STOMATA1

DOES manganese benefit plants mainly by increasing the oxidative power of the soil, as has been claimed by Skinner and Reid2 or is its chief value as a promoter of enzyme activity within the plant, as stated by Bertrand 78 McHargue has demonstrated

1 Contribution 354 of the R. I. Agricultural Experiment Station, Kingston, R. I.

2 Skinner, J. J., and Reid, F. R., "The Action of ManU. S. ganese under Acid and Neutral Soil Conditions."' D. A. Bull. 441. 1916.

8 Bertrand, Gabriel, "Sur l'intervention du Manganese dans les Oxidations provoqués par la laccase." Compt. Rend. Acad. Sci. (Paris) I: 124: 1032–1035.

4 McHargue, J. S., "The Rôle of Manganese in Plants.'' Jour. Am. Chem. Soc. 44: 1592-1594. 1922.

Since the need for manganese on neutralized soils appears to be so general with many kinds of plants it is worth while to know whether its action is mainly on the soil or within the plant itself. This question was answered in the experiment here described by. supplying some chlorotic plants with manganese through the soil, and introducing it into the tissues of others directly through the stomata of the leaves. This last was accomplished by an adaptation of the porometer, used by Darwin and Pertz' for studying stomata openings, and modified by McLean and Lees for inoculating citrus leaves with canker organisms. The apparatus consisted of a small glass medicinedropper tube with a rubber lip on the large end so that it could be pressed against a delicate leaf without causing injury. The small end of the tube was connected with a rubber atomizer bulb so that air could be forced into it under pressure. Then the tube was filled with a dilute manganese solution, its open large end pressed downward on a leaf, and the solution pumped into the intercellular spaces through the stomata. By using potted plants and tilting the pots on their sides, it was possible to inject the intercellular spaces of the leaves nearly full of the solution, then wash off with distilled water any surplus that might adhere to the leaves, without getting any of the solution into the soil.

In this way the effects were noted of supplying manganese to chlorotic spinach plants into the leaves through the stomata and also of supplying it to the

5 Gilbert, Basil E., McLean, Forman T., and Hardin, Leo J., "The Relation of Manganese and Iron to Limeinduced Chlorosis.'' Soil Science 22: 437-446. 1926. 6 Schreiner, Oswald, and Dawson, Paul R., "Manganese Deficiency in Soils and Fertilizers." Jour. Ind. and Eng. Chem. 19: 400-404. 1927.

7 Darwin, F., and Pertz, D. F. M., "A New Method of Estimating the Aperture of Stomata.” Proc. Royal Soc. London, Ser. B, No. B569: 136-154. 1911. Cited by Samuel F. Trelease and B. E. Livingston, "The Daily March of Transpiring Power as indicated by the Porometer and by Standardized Hygroscopic Paper." Jour. Ecol., No. 14: 1. 1916. Abstract in SCIENCE, New Ser., 43: 363. 1916.

8 McLean, Forman T., and Lee, H. Atherton, "Pressures required to Cause Stomatal Infection with the Citrus Canker Organisms." Philippine Jour. Sci. 20: 309-320. 1922.

soil. Equally prompt benefits were observed by both methods of treatment.

For this test six Wagner pots, each filled with about 10 kilograms of neutralized soil, were planted to spinach on April 20. On May 17, the plants had two to three leaves each and were very chlorotic. The pots were then arranged in pairs, each pair containing comparable plants. Then the plants in one of each pair of pots were treated with manganese sulphate solution. The treatments were as follows: Pot No. 79 150 cc. solution of 50 p.p.m. of manganese poured on the soil.

On May 24, one week after treatment, the plants injected with 50 p.p.m. manganese solution were greener than the control plants and showed the greatest improvement. The plants injected with 5 p.p.m. manganese solution and those receiving manganese through the soil were also greener than the control plants, but not equal to those receiving 50 p.p.m. On May 31, it was noted that the plants which received 50 p.p.m. of manganese were greener, but smaller, than those receiving only 5 p.p.m.

On June 7, the plants were harvested and weighed green, with the following results:

grams, and of the treated plants 6.6 grams, the average increase due to the manganese being 30 per cent. Manganese was apparently about equally effective whether injected into the tissues of the leaves or applied to the soil. Also, the control plants in Pot 15, which alternated with the injected plants in the same pot, were benefited neither in weight nor appearance by the treatment of the adjoining plants. So it is quite safe to conclude that this lime-induced chlorosis was cured by the action of the manganese within the body of the plant. The changes brought about in the soils by additions of manganese may be beneficial, but such changes were clearly not necessary for the recovery of the spinach in these experiments, while the injection of manganese solutions into the plants was clearly beneficial.

This method of injection of solutions into the leaf tissues through the stomata may be advantageously employed in the study of other diseases of plants suspected to be due to deficiency of soluble substances. FORMAN T. MCLEAN

RHODE ISLAND STATE COLLEGE

SOUTHWESTERN ARCHEOLOGICAL

CONFERENCE

ON August 29-31, 1927, there was held at the excavation camp of Phillips Academy, Andover, at Pecos, New Mexico, an informal gathering of workers in Southwestern archeology and related fields. There were present: C. Amsden, Southwest Museum; Monroe Amsden, Southwest Museum; Lansing Bloom, Museum of New Mexico; K. M. Chapman, Museum of New Mexico; H. S. Colton, University of Pennsylvania; C. B. Cosgrove, Peabody Museum of Harvard; Harriet Cosgrove; Byron Cummings, University of Arizona; A. E. Douglass, University of Arizona; Clara Lee Fraps, University of Arizona; Charlotte Gower, University of Chicago; O. S. Halseth, Arizona Museum; M. R. Harrington, Museum of the American Indian; E. L. Haury, University of Arizona; E. L. Hewett, Museum of New Mexico; Walter Hough, U. S. National Museum; N. M. Judd, U. S. National Museum, National Geographical Society; A. V. Kidder, Carnegie Institution and Phillips Academy; Madeleine A. Kidder; A. L. Kroeber, University of California; T. F. McIlwraith, University of Toronto; H. L. Mera, Indian Arts Fund; Paul Martin, Colorado State Museum; S. G. Morley, Carnegie Institution of Washington; Frances R. Morley; E. H. Morris, Carnegie Institution of Washington; Ann A. Morris; J. L. Nusbaum, National Park Service; Frank Pinkley, National Park Service; E. B. Renaud, University of Denver; Oliver Ricket

son, Carnegie Institution of Washington; Edith B. Ricketson; F. H. H. Roberts, Jr., Bureau of American Ethnology; Linda Roberts; J. A. B. Scherer, Southwest Museum; H. Shapiro, American Museum of Natural History; Leslie Spier, University of Oklahoma; Erna Gunther Spier; H. J. Spinden, Peabody Museum of Harvard; J. B. Thoburn, Oklahoma Historical Society; T. T. Waterman, University of Arizona; R. Wauchope, University of South Carolina.

The purposes of the meeting were: to bring about contacts between workers in the Southwestern field; to discuss fundamental problems of Southwestern history, and to formulate plans for coordinated attack upon them; to pool knowledge of facts and techniques, and to lay foundations for a unified system of nomenclature.

The morning of Monday, August 29, was spent in inspecting the academy's excavations in the pre-Pecos site at Bandelier Bend, and in visiting the main Pecos ruin. Monday afternoon and the mornings and afternoons of Tuesday and Wednesday were devoted to the business of the meeting, less formal campfire gatherings being held each evening. On Thursday, September 1, several members of the group visited the excavations of the School of American Research at Puyé by invitation of Director E. L. Hewett.

In the preliminary discussions, special attention was paid to the classification of Southwestern cultureperiods. There was entire unanimity in regard to the general nature of Southwestern culture-growth, i.e., that its basic element, maize agriculture, was derived from the South; that from time to time certain other highly important elements such as cottongrowing, loam-weaving, and probably pottery-making, were also introduced from the same source; but that little more than the germ-ideas of these elements penetrated to the Southwest; and that the development of its culture was essentially autochthonous.

There was practical unanimity as to the course of development, i.e., that agriculture was taken up by a previously resident, long-headed, nomadic or seminomadic people, who did not practice skull-deformation, and who already made excellent coiled basketry, twined-woven bags, sandals, and used the atlatl; but whose dwellings were of perishable nature. The newly acquired art of agriculture led to a more settled life and to the development of more permanent houses. For some time, however, pottery-making was unknown. At a later date pottery was introduced, or possibly independently invented, houses of the pit type were perfected, and became grouped into villages, and the bow-and-arrow began to supplant the atlatl. The long-headed race, however, still persisted,

At a still later period there appeared certain important changes: skull-deformation was initiated (the majority of those present at the conference believe that a new, broad-headed strain supplanted the ancient long-heads); dwellings emerged from the ground, the rooms became rectangular, and were grouped more closely; structural rings (corrugations) were for the first time left unobliterated on cooking vessels. From then on the development of the culture was rapid. After a period of wide extension, marked by small-village life, there was, perhaps a decrease in amount of territory occupied, and surely a concentration of population in certain areas, together with great architectural and ceramic achievement and strong regional specialization. Subsequently large areas were abandoned, there appears to have been a considerable shrinkage of population, and there was a definite cultural degeneration. This period was brought to a close by the settlement of the Southwest by the Spanish about 1600.

The meeting attempted, as a basis for more precise definition of culture-stages, to arrive at agreement as to diagnostic culture-traits. A sub-committee prepared a chronological tabulation of elements, which was used during the subsequent discussions. Architecture was considered to be of much value as an

index of growth; as were village-types, sandals, pictographs, etc. Much further information, both as to nature and distribution, was decided to be needed, however, before these categories can be used with full confidence. Pottery, it was agreed, is at the present time the most abundant, convenient and reliable criterion, and the cooking wares the simplest type for preliminary chronological determinations. Discussion brought out the following outline of development in this class of ceramics: first, plain wares; later, neck corrugations produced by leaving unobliterated the upper structural rings; still later, spiral corrugations ornamented by indentations and covering the entire vessel; again later, a degeneration of the corrugated technique, and, finally, disappearance of corrugations and return to plain-surface pots. During all the discussions leading to development of the above outlines, there kept arising questions of period nomenclature. Entire unanimity was not achieved, but the following terms for chronologically sequent periods proved acceptable to the majority:

Basket Maker I, or Early Basket Maker-a postulated

(and perhaps recently discovered) stage, pre-agricultural,

yet adumbrating later developments.

Basket Maker II, or Basket Maker-the agricultural, atlatl-using, non-pottery-making stage, as described in many publications.

Late Basket Maker, Basket Maker III, or Post-Basket Maker-the pit- or slab-house-building, pottery-making

stage (the three Basket Maker stages were characterized by a long-headed population, which did not practice skulldeformation).

Pueblo I, or Proto-Pueblo-the first stage during which cranial deformation was practiced, vessel neck corrugation was introduced, and villages composed of rectangular living-rooms of true masonry were developed (it was generally agreed that the term pre-Pueblo, hitherto sometimes applied to this period, should be discontinued).

Pueblo II-the stage marked by widespread geographi cal extension of life in small villages; corrugation, often of elaborate technique, extended over the whole surface of cooking vessels.

Pueblo III, or Great Period-the stage of large com. munities, great development of the arts, and growth of intensive local specialization.

Pueblo IV, or Proto-Historic-the stage characterized by contraction of area occupied; by the gradual disappearance of corrugated wares; and, in general, by decline from the preceding cultural peak.

Pueblo V, or Historic-the period from 1600 A. D. to the present.

As a by-product of the effort to define the various Pueblo periods, the following definition of a pueblo as an architectural type was arrived at: A pueblo is an agglomeration of essentially rectangular living rooms of adobe or masonry construction, generally flat-roofed and built above ground.

There was much discussion of the term "kiva" and of such parts of kivas as the ventilating passage, the fire-screen or deflector, etc. It was agreed that ceremonial rooms varied so greatly in form and in interior arrangement, and that the types shaded into each other so imperceptibly that no valid distinction as to essential function could be drawn between, for instance, round and square, or between above-ground and subterranean examples. The following very broad definition was therefore adopted: A kiva is a chamber specially constructed for ceremonial pur

poses.

It was hoped that the meeting could devote attention to the nomenclature of areas, of pottery types, pottery forms, elements of decoration, etc.; but so many matters of greater immediate interest were brought up that these questions were deferred with the idea that they should be kept in mind by those present and gone into at a possible future gathering. It was, however, thought well to consider the advisability of a binomial ware-nomenclature; the first name to be indicative of the locality of highest development, the second a technically descriptive term; for example, Sikyatki yellow, Mimbres black-onwhite, Upper Rio Grande incised, etc.

There was no opportunity for consideration of the difficult and at present very confused question of names for design-elements, but Mr. K. M. Chapman,