tains, where it is not needed for agriculture, it is deficient on the lowlands, where man has to irrigate. However, the aridity of parts of the West has some compensation in the extensive forests of tremendous trees on the soaking slopes of the Pacific. SEASONAL DISTRIBUTION OF RAINFALL In some respects, the distribution of rainfall throughout the year is more important than the amount. On this depends the rainfall usable for agriculture, and likewise the effects of rainfall on soil. Thus the 25 inches of rainfall in Nebraska are as useful for crops as 40 inches in Virginia. In fact, the extra 15 inches in Virginia may do more harm than good, on poorly kept farms at least, by washing and leaching the soil. Rainfall comes (1) in general cyclonic rains, (2) in local convectional (thunder) showers, and (3) in topographically produced falls. The cyclonic rains are greatest with frequent strong cyclones in regions where there is abundant moisture. The thundershowers are most numerous in mid-summer1 unless at this time the supply of moisture is not abundant. The topographically produced rains are heaviest when there is the greatest cooling of the moist winds. In the United States, general cyclonic rains on the Pacific coast and in the eastern third of the country are heaviest in the colder months. Thunderstorms are common in summer in the wetter parts of the country west of the SierraNevada-Cascades. Topographically produced rains are important on the Pacific, Gulf and Atlantic coasts, and on the windward sides of mountains; they are essentially early winter rains. Professor Ward has picked out 14 wellrecognizable rainfall types in the United States; and he has made composite curves and discussed each. The rains east of the Rockies tend to be heaviest in summer, and those west, in winter. The type covering the greatest region is the continental "Missouri" type. It 4 See Mo. Weather Rev., Vol. 43, 1915, pp. 322340, 13 charts; and pp. 619–620. 5 Geogr. Review, Vol. 4, 1917, pp. 131-144. has a summer rainy season with a maximum of over 4 inches in June and a minimum of 1 inch in January. This shades off into many types on all sides. The Ohio type may be considered as the Missouri type with 1 to 2 inches of cyclonic rain added through the cold half of the year. The New England type has still more of the cyclonic winter rainfall, with 3 to 4 inches of rain every month. Farther south the Atlantic has an intensification of the July and August rainfall with the very favorable moisture conditions for thunderstorms and with the occasional heavy rain of tropical cyclones. The Tennessee type includes so much rainfall from the strong cyclonic action in February and March that the principal maximum, 4 inches comes at this time; and there still is the summer maximum. The Gulf coast is always moist. There are three types of rainfall-different combinations of thunderstorm and cyclonic rains—all with maximum intensity in the warmer half of the year. In the East Rocky Mountain Foothills type, the rainfall in spring starts off like the Missouri type, but the winter snows are insufficient to supply moisture for increasing thunderstorm rains beyond May. The winters are dry in spite of numerous cyclones, because the air can contain so little moisture at the low temperatures. West of the crest of the Rockies, the moisture from the Pacific is precipitated topographically most in winter. In the plateau region, summer convection, especially before the ground is thoroughly dried, brings another maximum early in summer. In the south, however, the winter precipitation is so light and so soon evaporated that the şummer showers do not occur till July when moisture arrives in sufficient quantity from the Gulf of California and the Pacific. On the north Pacific coast where there is much cyclonic activity throughout the winter the maximum comes in December (over 7 inches) when the topographic rainfall tends to be heaviest. In the south, cyclonic activity is more important than the cooling of on-shore winds in producing rainfall, so the heaviest rains in the "Southern Pacific" type occur from January SPECIAL ARTICLES A PARALLEL MUTATION IN DROSOPHILA A MUTANT of Drosophila funebris Fabr. has recently appeared that is so strikingly similar to a well-known mutant of D. melanogaster Meig. (ampelophila Loew) that there can be litle doubt that the same mutation has occurred independently in the two species. The new form, called notch, agrees with the notch melanogaster in at least eight different respects, as will appear below. Origin. A female funebris of a stock from Mitchell, S. D., was mated to a male of a stock from New York City. The descendants were mated in pairs for several generations, and no variations were observed except an occasional fly with one of the anterior scutellar bristles missing. Such flies were found also in the uncrossed New York stock. In the line under consideration selection was carried on, in an attempt to increase the percentage of such flies, but no marked result was obtained. In F, one pair (5201) produced 35 normal females, 34 notch females, and 36 normal males. The sex ratio here is significant, since an excess of males is more frequent than an excess of females in this species. The pair from which the parents of 5201 came produced 19 females and 31 males, which is not an unusual excess when complete counts are not obtained. In D. funebris the males usually emerge in a little less time than the females. This relation is just the reverse of that found in D. melanogaster. Evidently the female parent of 5201 was genetically notch. She was not observed to be abnormal, and had been destroyed when her offspring began to emerge. It seems probable that she did not have notched wings, but she may well have had the characteristic veins and acrostichal hairs, since these would more easily have been overlooked. Description.-Notch melanogaster is characterized by having the wings somewhat nicked, more especially at the apical posterior corner. But this character is somewhat variable, being often unlike in the two wings of the same female, and sometimes even entirely absent.1 In addition the eyes are often smaller than those of the wild-type flies and somewhat roughened.2 Furthermore the veins of notch are somewhat thickened, more especially the apical portions of the second and fifth longitudinal veins. This character is the most invariable and convenient index of the presence of the notch gene. The anterior scutellar bristles of notch are often doubled. The acrostichal hairs are more numerous than those of the wildtype fly, and are irregularly arranged, instead of being in eight fairly definite rows.3 The notch gene thus produces an unusually large number of morphological peculiarities. Notch funebris agrees in all of the above respects. The wings are nicked in the same way, but are often asymmetrical and sometimes normal; the eyes are often small and roughened; the wing veins are thickened even more than those of notch melanogaster, the second and fifth being affected most, and this character being again the most convenient and reliable for purposes of classification; the anterior scutellar bristles are often doubled, in spite of the fact that notch arose in a family selected for the absence of these bristles; the acrostichal hairs are irregularly ar 1 See Morgan, 1917, "The Theory of the Gene," Amer. Nat., 51, for figure and a discussion of this variability. 2 Bridges has shown that notch is probably an allelomorph of the roughened eye known as facet. Metz and Bridges, 1917, "Incompatibility of Mutant Races in Drosophila," Proc. Nat. Acad. Sci., 3. 3 The peculiarity of the acrostichal hairs was not observed here until it was looked for after notch funebris was found to have unusual acrostichals. ranged, but differ from those of notch melanogaster in being entirely wanting on each side in a narrow band just inside the dorsocentral row. The unusual features of notch in melanogaster are not limited to its morphological nature. Notch is one of the few dominant mutant genes, and in addition is sex-linked and has a recessive lethal effect. The result is that a notch female gives equal numbers of wild-type and notch daughters and of wildtype sons. Notch males never appear. This is the only known dominant sex-linked gene that is also lethal-except funebris notch. We have seen that the original notch culture, 5201, gave the characteristics 1: 1: 1: 0 ratio; and this has been repeated by the notch females produced in that culture, both when mated to their brothers and when mated to unrelated wild-type males.* The striking parallel between these two mutants makes it highly probable that they represent the same genetic change. This view is strengthened by the fact that notch is one of the most frequent mutations in melanogaster (known to have occurred seven times), and might therefore be expected to be one likely to occur in another species. Summary.-Notch melanogaster and notch funebris agree in the following respects: 1. Wings usually irregularly nicked at tip. 2. Certain veins thickened. 3. Eyes often small and roughened. 4. Acrostichal hairs not in definite rows. 7. Gene has a recessive lethal effect. 8. Gene is sex-linked in melanogaster, almost certainly so in funebris. 9. Mutation is one of the most frequent in melanogaster, and the first certain one in funebris. A. H. STURTEVANT COLUMBIA UNIVERSITY, May, 1918 It is theoretically possible that funebris notch is not sex-linked, but that the gene is dominant in females, lethal in males. This can be determined by finding gynandromorphs, or by finding other sex-linked genes and observing their linkage to notch. THE KENTUCKY ACADEMY OF SCIENCE THE Kentucky Academy of Science held its fifth annual meeting at the University of Kentucky on Saturday, May 4, 1918, with Mr. J. E. Barton, vice-president, in the chair. After a brief business session, at which several new members were elected, the following program was presented: President's address, by J. E. Barton, acting president, "The regenerative forests of eastern Kentucky and their relation to the coal-mining industry." The extensive coal-measures of eastern Kentucky support a valuable forest growth, which is of great usefulness in the mining of coal. At the present time it takes about three acres of timber to mine one acre of coal. The ratio should be nearly one acre of timber to one acre of coal. This condition can be brought about by careful management, which is justified by the fact that the coal supply will last about one hundred years, at present rate of production. Timber can be raised in a thirty-year rotation, of sufficient size and character for mining purposes, by a proper selection of species, an area fully stocked and adequate protection against fire and live stock. Differences in the ossification of the male and female skeleton: DR. J. W. PRYOR. Scientific education: J. J. TIGERT. The rapid development of scientific agriculture. Education followed agriculture in scientific progress. Scientific procedure dependent upon quantitative measurement. Statistical methods and measurements in education. Standard tests. The measurement of intelligence. Charts and tables showing results of measurements in the Cynthiana schools in 1916-17 and the Lexington schools in 1917-18. Age-grade table, Cynthiana, shows 22 per cent. of pupils retarded. Comparison of promotions in Cynthiana and other American cities shows a larger percentage of promotion in Cynthiana than elsewhere. Ayres Spelling Test in Lexington and Cynthiana shows Lexington three points above the average of 84 American cities, and Cynthiana equal to the average of 84 American cities. Handwriting tests in Lexington and Cynthiana show both these cities below the average city in Arithmetic speed and quality of handwriting. tests in Cynthiana show Cynthiana below standard measured by the Woody Scale. A comparison of boys and girls in spelling and handwriting shows the girls to be superior to the boys. The effect of manganese on the growth of wheat: J. S. McHARGUE. After reviewing briefly some noteworthy results obtained by previous investi gators on the relation of manganese to agriculture, the author presented results obtained by growing wheat in manganese-free sand and in cultural solutions, with and without the addition of manga nese. Wheat plants grown to within a few weeks of maturity in cultural solutions containing manganese and others of the same age in which the manganese had been omitted, were on exhibition. Where manganese had been added to the cultural solutions the plants were apparently normal in every respect, whereas the plants grown in solutions containing no manganese showed a retarded growth in the blades, stalks and roots, as compared with the plants of the same age receiving manganese. There was evidence of lack of the proper development of chlorophyl in the plants receiving no manganese and the blades of these plants exhibited a drooping appearance in that they were not able to hold themselves erect, which was quite characteristic and not to be observed in any of the plants receiving manganese. The author concludes from his experiments that manganese plays a more important role in the growth of wheat than has hitherto been suspected. Formation of petroleum: C. J. NORWOOD. (By title.) Cryoscopic work with an ordinary thermometer: C. C. KIPLINGER. It has been found possible to read small temperature intervals on a common thermometer, within an accuracy of 1/100 degree, by measurements of the parallax on an auxiliary scale equipped with a sliding peep-sight. Several heretofore troublesome sources of error in the boiling point method of determining molecular weights have been eliminated by using but one point as reference on a thermometer scale, having established this point by the use of a known substance with a high degree of purity. This procedure eliminates the need of a calibrated thermometer. The use of the parallax method is suggested in the estimation of fractional parts of a scale division on other instruments than the thermometer. Generalization on the mean-value theorem: H. H. DOWNING. Magnolia fraseri: does it occur in Kentucky? FRANK T. MCFARLAND. List of fungi from Kentucky: FRANK T. McFARLAND. An equation balance: E. L. REES. A method of constructing the graph of an equation in which the variables may be separated: E. L. REES. Protein metabolism in the growing chick: G. D. BUCKNER and others. (By title.) Review and observations on the mosaic disease of tobacco: G. C. ROUTT. The author reviews the work of other investigators and reports observations of his own upon the disease in experimental plots of different varieties of tobacco. He favors the view that the best way to combat the disease will be to develop a resistant strain of tobacco. Dr. J. A. Detlefsen, of the department of genetics of the University of Illinois, addressed the academy on "Laws governing the transmission of characters from parent to offspring." The speaker gave a brief review of the search by investigators for the cause or causes of evolution. He then explained the law for the transmission of mono-hybrids, di-hybrids and tri-hybrids. He presented these laws and illustrated them so well that there was left no doubt in the minds of workers in other fields that great progress has been made in genetics in recent years. He threw upon the screen the tables giving the result of his own breeding experiments to show how nearly actual counts agree with the mathematical expectation, in the laws of transmission. It is remarkable how nearly actual counts of animals bred agree with the expectation of what, by Mendel's law, they should be. Among other items of business, a resolution was passed offering the services of the academy to the U. S. government for any war work in which this organization might be of assistance. Officers were elected as follows: J. E. Barton, Frankfort, President; P. P. Boyd, Lexington, Vice-president; A. M. Peter, Lexington, Secretary; J. S. McHargue, Lexington, Treas SCIENCE FRIDAY, JULY 26, 1918 AMERICAN ASSOCIATION FOR THE ADVANCEMENT OF SCIENCE COMMITTEE OF ONE HUNDRED FUNDS FOR RESEARCH IN ASTRONOMY WHEN drawing up the report upon Research Funds made to the American Association in December, 1915, and subsequently printed in SCIENCE there seemed to the committee to be good reason for believing that it would be advisable to place the data relative to astronomical observatories in a separate article, together with certain additional facts which would be of value to those particularly interested in astronomical research. For this reason a circular letter of inquiry, dated February 1, 1917, was sent to the principal American observatories asking a reply to the following questions in each case: 1. What are the principal and annual interest of observatory funds available for research as distinguished from teaching and what fraction of the income as far as can be estimated may be credited to research? 2. What are the stated publications of the observatory or other papers indicating the results of researches accomplished? The replies to this letter are uniformly clear and full. Abstracts of them are given below with data taken in some cases from official publications. It is thought that this form of presentation is preferable to a mere tabulation inasmuch as a more definite idea may thereby be secured as to the conditions which obtain in each of the observatories concerned. Especially does this seem desirable in that those interested in astronomy though not professionally engaged in its pursuit may find a brief but intelligible statement of what provision has been made in this country up to the present for the actual advancement of the science by research. The undersigned will be glad to receive cor |