SCIENCE. 



[Vol. XI. No 265 



possible through an increase of rainfall, such increase must be of 

 notable amount. The effect upon agriculture of a minute increase 

 would be scarcely appreciable, and certainly would not suffice to 

 produce the effects claimed for it, or to explain the wide-spread 

 belief in this increase which is prevalent. In examining the rain- 

 fall records, we are, then, to look for substantial amounts of in- 

 crease, — several inches annually. I would add that these records 

 are now ample for testing this theory, and their testimony should 

 be conclusive. 



I find in this area twenty-six stations at which rainfall records 

 have been kept for periods ranging from six to twenty-eight years, 

 the total number of years of record being three hundred and ten- 

 These stations are scattered widely over the area in question, from 

 its eastern to its western border, and involve all stages of settle- 

 ment. Now, if there has occurred an increase in the amount of 

 rainfall, that of the later years of any series should, on the whole, 

 be greater than that of the earlier years. I have therefore cut each 

 of these series in the middle, and added up the rainfall of each half. 

 These are presented in the following table, where the first column 

 gives the names of the stations ; the second, the number of years 

 in the series ; the third and fourth, the total rainfall in the first and 

 second halves of each series respectively ; and the fifth, the increase 

 or decrease, the former being distinguished by the +, the latter by 

 the — sign : — 



Fort Leavenworth, Kan. 



Leavenworth, Kan 



Manhattan, Kan 



Ldwrence, Kan 



Fort Lamed, Kan 



Topeka, Kan 



Dodge City, Kan 



Wallace, Kan 



Atchison, Kan 



Baxter Springs, Kan 



Burlingame, Kan 



Council Grove, Kan 



Fort Hays, Kan 



Fort Riley, Kan 



Olathe, Kan 



Belleville, Kan 



DeSoto, Neb 



Fort McPherson, Neb. . . 



North Platte, Neb 



Omaha, Neb 



Omaha Agency, Neb . . . 



Yankton, Dak 



Bismarck, Dak 



Fort Benson, Mont 



Cheyenne, Wyo 



Denver, Col 



28 



51S 



525 



18 



366 



362 



28 



400 



407 



18 



306 



319 



.2 



131 



119 



8 



117 



140 



12 



los 



149 



6 



50 



59 



8 



189 



156 



6 



130 



102 



6 



84 



96 



8 



.78 



141 



6 



55 



79 



16 



185 



214 



8 



201 



194 



14 



184 



218 



6 



109 



80 



6 



58 



52 



12 



108 



120 



iS 



319 



337 



6 



75 



78 



12 



170 



178 



12 



140 



102 



6 



34 



40 



j6 



84 



98 



14 



112 



103 



+ 7 

 - 4 



+ 7 

 + 13 



- 37 

 + =4 

 + 29 

 — 7 

 + 34 



+ 18 

 + 3 



It will be seen at once that the individual results are contradic- 

 tory in a high degree ; those from sixteen stations showing an in- 

 crease, while ten stations show a decrease. These contradictions, 

 which are due to the irregularity of the rainfall may, however, be 

 in a measure eliminated by combining the results, under the sup- 

 position that the change, if any, has been a progressive one. Un- 

 der this assumption, the sum of the earlier halves of the different 

 series should be less than that of the later halves. Adding them 

 together, it is found that the aggregate rainfall at all the stations 

 was, in the first half of the series, 4,408 inches, and in the second 

 half, 4,468 inches ; showing that there has apparently taken place 

 an increase of 60 inches in the total amount of rainfall at all these 

 stations in a total of 310 years, or, to put it in another form, there 

 has fallen in each year of the second half of these series 0.4 of an 

 inch more rain than in the first half. It is unnecessary to add that 



this is not the sort of increase for which we were searching, as an 

 increase of but a fraction of an inch certainly could not produce 

 the results which are claimed. An examination of the seasonal 

 distribution of the rainfall shows that that also has undergone no 

 material change since settlement began in this region. We may 

 therefore dismiss as baseless the popular idea of an increase in 

 rainfall, either annual or during the growing season, and look else- 

 where for an explanation of the phenomena of settlement which 

 the plains preseijt. 



The early explorers, of the time of Fremont and the Pacific Rail- 

 road surveys, based their judgments of the capabilities of the 

 country for agriculture upon the character of the natural products, 

 the absence of trees, the presence only of sparse, hardy grasses, the 

 cactus, and the yucca. Their judgment was a mistaken one, as 

 events have amply proved. 



Since their time physical geographers have set arbitrary limits to 

 safe farming without irrigation, basing their reasoning upon the 

 known rainfall of the region, and that supposed to be required for 

 the average farm product. Subsequent experience has shown that 

 a much smaller quantity of rain is essential than was supposed. 

 To my mind, there is little more to be said. If it be found, that, 

 with an annual rainfall during the growing season not greater than 

 ten inches, farming can be carried on successfully, the only question 

 remaining is, how the mistake could have been made of supposing 

 that it required a greater amount. 



There is no doubt that cultivation adds greatly to the economy 

 of the rainfall. The surface of the plains in an uncultivated con- 

 dition is mainly bare, hard ground, but slightly protected by its 

 covering of grasses. From such a surface the rain flows off freely, 

 and an unusually large proportion of it finds its way into the 

 streams, while a correspondingly small proportion sinks into the 

 ground. The farmer, with plough and harrow, changes all this, 

 and retains in the soil most of the rainfall. From year to year the 

 supply in the soil increases, so that the subsoil becomes in time a 

 reservoir from which the surface soil may draw in times of drought. 

 Furthermore, the scanty vegetation offers little protection against 

 evaporation, which is excessive upon the barren plains ; but the 

 ampler mantle which cultivation spreads over the soil prevents its 

 moisture from disappearing in the atmosphere with so great 

 rapidity. 



How much farther westward into the arid region can the farmer 

 push ? This is a very important question, affecting the value of 

 millions of acres of land ; for, if this land can be cultivated only by 

 the aid of irrigation, nine-tenths of eastern Montana, Wyoming, 

 Colorado, and New Mexico, together with western Dakota, Ne- 

 braska, and Texas, must be given over to the cattle-men in per- 

 petuity, as the streams are entirely insufficient for irrigation. A 

 conclusive and satisfactory answer can be given only by the farmer 



Henry Gannett. 



WASHINGTON SCIENTIFIC NEWS. 

 A Novel Way of Forming Alloys. — The Constituents of Sugar. — 

 Rainfall in the Arid Regions. — Irish Myths and Folk-Tales. — 

 Examining Fats. 



The Formation of Alloys. 



The following is an abstract of a note read before the Philo- 

 sophical Society by William Hallock, of the United States Geologi- 

 cal Survey, Feb. 18, 1888 : — 



In the Berichte der chemischen Geselhchaft,\o\.-^\. 1S82, pp. 

 595-597, W. Spring describes the formation of alloys by submitting 

 the filings of the constituent metals to high pressure, without ap- 

 preciable rise in temperature. Wood's alloy of cadmium, tin, lead, 

 and bismuth he produced by mixing proper weights of the filings 

 of these metals, and subjecting them to a pressure of 7,000 at- 

 mospheres. The block thus obtained was again filed up, and 

 subjected to the same pressure. 



In this way a block of metal was produced which possessed the 

 physical properties of ordinary Wood's alloy, formed by melting the 

 mixed constituents. 



W. Chandler Roberts repeated this experiment {Chemical News, 

 vol. xlv. 1882, p. 231), and verified Mr. Spring's results. 



In seeking an explanation of the above phenomenon satisfactory 



