land surface is not continued beneath the Cretaceous formation of the Coastal Plain. The much greater slope of the crystalline basement below the Coastal Plain makes the marked angle formed by the intersection of these two surfaces recognizable even when a small vertical exaggeration is used. When records of scattered well drillings along a line in the Coastal Plain are plotted, they seem to indicate that the crystalline basement closely approximates a plane surface. Very probably the surface of the crystalline basement is a buried peneplane developed during pre-Cretacic (Jurassic?) times. The slope of this buried and tilted Jurassic peneplane emerges from beneath the Coastal Plain sediments, and continues upward along the face of the fall line belt of the Piedmont, being revealed through the stripping away of the Coastal Plain sediments by erosion. The width of this stripped zone is slight, varying from 2 to 4 miles along the Piedmont. In New England, however, the Coastal Plain is entirely removed (except that portion which is below sealevel), leaving a stripped zone varying from 5 to 15 miles in width. The exposed edge of this old Jurassic peneplain is continued across the bottom of the Gulf of Maine with a width of about 15 miles. Several writers have long known that the slope of the Piedmont is not continued beneath the Coastal Plain and that there are two upland slopes in New England; indeed, Davis himself has recognized this and commented on it. But what is vastly more significant, it has not been recognized that since this is true no falls could result from the conditions embodied in Davis's theory. If, as Davis suggested, the Piedmont rivers had established very gentle gradients leading to an ancient shore-line near the present "fall line," the upland profile and the stream profile must have been intersected at that point. After the emergence of the Coastal Plain, the extended rivers must have entrenched themselves in the soft Coastal Plain sediments until they had established nearly level channels in their lower courses, whereas downcutting proceeded very slowly in the crystalline rocks of the Piedmont. This would have resulted in a broken stream profile in which the stream in the outer (eastern) Piedmont would have been entrenched but little below the upland surface, while in the western Piedmont and the Coastal Plain it would have been greatly entrenched. But actual profiles show that the streams are entrenched as far below the upland surface in the eastern Piedmont as in the western Piedmont. If this fact is applied to a diagram showing the Piedmont surface as being continued beneath the Coastal Plain it gives only a simple concave stream profile which the postulate of an ancient shore-line at the fall line is untenable, and the break in stream gradient is only incidentally related to differential entrenchment in areas of unequal rock resistance. On the other hand, the profiles show that there is a marked break in the slope of the Piedmont-New England upland, and it is at this point that the streams which are uniformly entrenched below the upland surface exhibit the break in their gradients. The fall line zone, therefore, lies along the outer edge of the crystalline area (the Piedmont-New England Upland Province), where there is a break in the slope of the land due to the intersection of the recently exposed margin of the old, tilted Jurassic peneplain and the newer Tertiary peneplain of the uplands. Since the gradients of the streams are closely related to the profile of the land, this break is sufficient to account for falls and rapids. GEORGE T. RENNER UNIVERSITY OF WASHINGTON SCIENTIFIC APPARATUS AND LABORATORY METHODS PRECIPITATION OF THE VIRUS OF TOBACCO MOSAIC IN an attempt to free the virus of tobacco mosaic from as much contaminating material as possible a method has been devised whereby the freshly cut diseased tobacco tissues are frozen, allowed to thaw, and then subjected to high pressure. The juice obtained, after being centrifuged, at about 2000 r.p.m., contains no large particles in suspension, but is highly infectious. When two volumes of acetone (c. p.) at -8° C. are added to one volume of the juice held at about 0° C. a flocculent precipitate is thrown out and rapidly settles. The supernatant liquid can be almost completely decanted within two minutes after adding the acetone, leaving the precipitate in the bottom of the container. More water may then be removed by rinsing the precipitate with acetone (c. p.) at -8° C., decanting the acetone and then removing the remaining acetone with absolute ether at -8° C. by rinsing twice and thoroughly draining off the ether. The precipitate thus obtained is readily soluble in distilled water. Experiments in which young tobacco plants were inoculated with this solution showed it to be highly infectious. The first supernatant liquid decanted, on the other hand, when centrifuged to free it from all traces of the precipitate and diluted with two parts of distilled water proved to be non-infectious. The dilution was made in order to bring the concentration of acetone below the point which Allard1 showed was non-toxic to the virus. Absolute alcohol may be used in place of acetone under the above-mentioned conditions. At about 100 per cent. saturation and -8° C. ammonium sulfate salts out from the juice, material which, when filtered off and sucked dry, dissolves readily in distilled water. Plants when inoculated with this solution take the disease. The filtrate when diluted, one to five, has in no case transmitted the disease; although the untreated juice when containing ammonium sulfate solution at a concentration of 3 cc. of a saturated solution to 10 cc. of the juice is infectious. Solutions of Safranin-O have also been used to precipitate the virus from the plant juice. This gives a quantitative precipitation, which frees the juice of virus. BOYCE THOMPSON INSTITUTE FOR PLANT RESEARCH, YONKERS, NEW YORK CARL G. VINSON BROWNIAN MOVEMENTS WITH LOW THE desire having arisen for conspicuous Brownian movements, a variety of materials was pulverized with a view to ascertaining which showed the movements to best advantage. For several reasons mica, particularly in the form of muscovite, was found preferable. The suspension to be observed may be prepared as follows. A quantity of mica from the edge of a natural slab is ground by a dry emery wheel into an impalpable dust. This is stirred into a graduate of water and allowed to stand for some four hours. After the larger flakes have settled to the bottom, the thin milky suspension is siphoned off, care being taken not to draw off any of the useless residue at the same time. The concentration of the liquid may, of course, be altered as seems convenient by evaporation or dilution. The liquid so prepared contains particles most of which are so small as to exhibit the Brownian movements. Under a magnification of fifty diameters with oblique illumination from below the microscope stage, the flakes appear as bright scintillating points in a dark field. This scintillation is evidently caused by small angular displacements due to the atomic bombardment; as the flakes rotate, they reflect the light at irregular intervals. Mica is peculiarly well adapted to this method of observation because each thin particle has a moment of inertia small in comparison with its reflecting area. 1 Allard, H. A. Jour. Agr. Res. 13: p. 619 (1918). HYSTERESIS LOSS IN NICKEL OF THE hysteresis loss in specimens of nickel crystals, which varied from one grain per specimen to as high as 2.8 × 10o, has been examined by Sucksmith and Potter1 and found to increase rapidly as the number of crystal grains increased. Such an effect is not limited to specimens specially prepared as crystals. but may be found as well in strips of ordinary nickel which have been successively cold rolled to thinner and thinner specimens and thus the number of crystals per unit volume increased step by step as well as the hardness. Of course all metals are crystalline, but in the case of the nickel strips which are cold rolled, it is not until severe cold working is performed that the crys tals are more or less aligned2 in one direction. The present writer had occasion recently to study some of the magnetic properties of a series of eleven nickel strips reduced to various thicknesses by suecessive cold rolling. These strips were 57.7 cm. long and about 0.954 cm. wide. The thickest strip was .604 cm. in thickness and the ten succeeding strips were rolled from this thickness to those given by the percentage cold reduction in the following table: and hysteresis loss. This is shown in Fig. 1 and confirms the relation found by others that large grain size is conducive to small hysteresis loss. Honda and Kaya' in a recent study of the magnetic properties of single crystals of iron find a similar law holding in their work. The same conclusion may be drawn from Gerlach's curves for the magnetization of single iron crystals and electrolytic specimens. Sorensen ascribed the high coercive force in thin films of iron, cobalt and nickel as due to the minute size of the crystals. Edwards' had a similar experience. The recent work of Ishagakis on the effect of grain-size on the hardness of pure iron fits into the same picture. It is interesting to note, on the other hand, that Welo and Baudisch' found for precipitated magnetite that "lean hysteresis loops, low coercitivities and low remanences are associated with oxides composed of small crystals." S. R. WILLIAMS FAYERWEATHER LABORATORY OF PHYSICS, AMHERST COLLEGE, AMHERST, MASS. 3 Williams, Proc. A. S. S. T., 1926. Honda and Kaya, Sci. Reps. Tohoku Imp. Univ., 15, p. 729, Nov., 1926. Gerlach, Ztschr. f. Phys., 38, p. 832, 1926. * Sorensen, Amer. Phys. Soc. Program, Abstract, Nov. 28-29, 1924, Phys. Rev., 24, p. 658, 1924. Edwards, Amer. Phys. Soc. Program Abstract, Dec. 28-30, 1925. * Ishagaki, Sci. Reps., Tohoku Imp. Univ., 16, p. 285, 1927. Welo and Baudisch, Amer. Phys. Program, Abstract, Feb. 26-27, 1926. A DYSENTERY-LIKE BACILLUS FROM A PHLEGMONOUS INFLAMMATION THE bacilli belonging to the dysentery group have with few exceptions been isolated from the intestinal and urinary tracts. The writer has found but one reference to the isolation of one of this group from an extremity. Magnusson, 1919,1 isolated a dysentery bacillus, which he named Bacterium viscosum equi, from "joint ill" in foals. Since the bacillus to be described was isolated from a phlegmonous inflammation of the lower leg and foot of a man, it will, perhaps, be of interest. The isolated bacillus has the following characteristics: Non-motile, Gram-negative, non-spore-forming, short were first brought into prominence by Elizabeth Sidney Semmens (Journ. Soc. Chem. Ind., 42, 954, 1923, also Brit. Assoc. Report, 1923). In two previous notes by Bhatnagar and Lall and Bhatnagar, Lall and Mathur (Nature; February 27, 1926, and July 3, 1926) the effect of polarized light with the electric vector in the light-wave vibrating in the plane of incidence was shown to be selective on animal metabolism and on growth of V. cholerae and B. typhosis. Further work on the subject has been published from this laboratory in the Indian Journal of Medical Research (Vol. XIV, No. 2, October, 1926). No positive results seem to have been so far recorded in favor of a selective effect on purely photochemical reaction, though there are some unsuccessful attempts described in literature (Ghosh, Journ. Ind. Chem. Soc., 1925, Vol. 2, p. 269). Investigations were therefore undertaken to find whether polarized light would accelerate purely chemical reactions as it does some of the biochemical reactions studied by Baly and Semmens, Hill and Macht and Bhatnagar, Lall and Mathur. Positive results have now been obtained in the case of the interactions of liquid amalgam of sodium or potassium and water. This reaction was shown to be photo-sensitive by Bhatnagar, Mata Prasad and D. M. Mukherjee (Journ. Ind. Chem. Soc., 1925, Vol. 1, 263). The apparatus employed to obtain two fairly large patches of the polarized light and ordinary light of the same intensity was the same as described in a previous note (Nature, July 3, 1926). Spectra of the two beams of light taken in the visible region by means of an Adam Hilgers Spectrograph were found to be identical. The heat-rays were cut off from the reaction vessels by interposing in each arm of the apparatus a rectangular glass cell containing a strong solution of alum. The equality of intensities was measured by means of a Hilger Thermopile and Broca Galvanometer as described in the note referred to above. The polarized beam indicated a polarization of 90.5 per cent. as measured by means of a Savart's polariscope. The reaction between the amalgams of the alkalimetals and water takes place in the dark, but is considerably accelerated by light even in the visible region. As a result of this reaction hydrogen and sodium hydroxide are produced. The rate of the reaction was studied in two ways: (1) By measuring in a capillary tube the movements of a column of mercury due to the generation of hydrogen, (2) by titrating the alkali produced against a standard solution of an acid. Both sets of experiments showed a remarkable acceleration of reaction when the reaction vessels were exposed to polarized radiations with the electric vector vibrating parallel to the plane of incidence. These results have been verified in hundreds of experiments. The differences in accelerations produced are not small and with suitable surfaces of the alloys exposed to the two kinds of radiations, they are as large as thirty-three per cent., and a complete account of the work is being communicated to one of the chemical journals. As far as the author knows, this is the first purely chemical reaction as distinguished from the biochemical reactions studied by previous investigators which has definitely been shown to be selectively affected by polarized radiations. When the radiations polarized with the electric vector vibrating perpendicular to the plane of incidence, fall on the reaction vessels. minimum accelerations are produced in the reaction. Work on this subject, however, is in progress and definite results will be communicated later on. A large number of other photochemical reactions such as the photo-decomposition of hydrogen perox ide, the reaction between mercuric chloride and ammonium oxalate in presence of iron chloride have been tried with negative results. From a large number of the photochemical reactions studied from this standpoint, it appears so far that the photochemical reaction in heterogeneous system only show this selective action and that the surface plays some part in this reaction. In this connection it is interesting to recall the remarkable discovery which Elster and Geitel made in 1894 that in certain cases the photoelectric effect is influenced by the orientation of the plane of polarization of the incident light. Using also the liquid alloy of sodium and potassium and allowing polarized white light to fall upon surface, at an angle of 45°, they found a maximum current when the electric vector in the light wave was vibrating in the plane of incidence and a minimum current when the electricvector was perpendicular to the plane of incidence. These experiments have been confirmed by Kunz and a number of other workers and have received a satisfactory explanation through the work of Pohl and Pringshein (Deutsch. Phys. Gesell, Veb., 12, p. 215228, 349, 360, 1910). It is proposed to apply and extend the view of Pohl and Pringsheim regarding this selective effect of polarized radiations to the case of photo-chemical reactions described in this note. S. S. BHATNAGAR UNIVERSITY CHEMICAL LABORATORIES, UNIVERSITY OF THE PUNJAB, LAHORE |