140 



THE POPULAR SCIENCE MONTHLY. 



cold current is of least strength, as in Au- 

 gust and September, the Gulf Stream comes 

 within ten miles of Barnegat ; at other times 

 it is distant one hundred and twenty miles, 

 changing with the amount of the cold cur- 

 rent and of the wind. If we had not the 

 cold wall between our shores and the Gulf 

 Stream, it is fair to presume that we should 

 have a less stormy coast, as the juxtaposi- 

 tion of these two currents with their differ- 

 ence in temperature must from that circum- 

 stance tend to an unstable condition of at- 

 mospheric equilibrium. Our cold northwest 

 winds would then sweep to the north of 

 us, and become westerly and southwesterly 

 winds. 



Production of Coke in the United States. 



It is shown by the report of Mr. Joseph 

 D. Weeks, of Pittsburg, to the United 

 States Geological Survey, covering the pe- 

 riod from 1880 to 1885, inclusive, that 

 Pennsylvania stands first in the rank of 

 coke - producing States, Alabama second, 

 West Virginia third, and Tennessee fourth. 

 The largest coke-producing locality in the 

 country is the Connellsville region of Penn- 

 sylvania, in which were made 3,096,012 of 

 the 5,106,696 tons, or 60'6 per cent of the 

 coke produced in the United States in 1885. 

 The second largest producing district is 

 what is called the Irwin-Latrobe district, 

 which lies along the Pennsylvania Railroad, 

 from Larimer to Blairsville, and is, in part, 

 the northerly extension of the Connellsville 

 coking-field. The number of establishments 

 has slightly decreased, but the number of 

 ovens was increased from 1S84 to 1885 by 

 2 - 8 per cent, and was in the latter year 

 20,116. While the production of 1885 in- 

 creased over that of 1884, it was not as 

 great as in 1883. There has been no in- 

 crease in the value per ton of coke for three 

 years. 



How Ilarbor - Channels may be kept 

 clear. Professor Lewis M. Haupt addressed 

 the Section of Mechanical Science and En- 

 gineering of the American Association on 

 " River and Harbor Improvements, with 

 Special Reference to the New York En- 

 trance." He maintained that large and 

 weighty structures intended to regulate cur- 

 rents, which rest or depend upon sandy or 



alluvial bottoms, violate the fundamental 

 requirements that they shall not oppose the 

 ingress of the tide or injuriously modify the 

 currents. Dikes and jetties also are below 

 the plane of action of waves of translation ; 

 while dependent upon their mass they are 

 not entirely coherent ; and they are wasteful, 

 and result in serious modifications in the 

 regimen of rivers and harbors. He suggested 

 as a preferable system, one consisting of 

 deflectors intended to be attached to buoys 

 or floats, anchored to heavy moorings, and 

 guyed in place on the ebb side by wire 

 cables or chains. This system is composed 

 of units or parts readily assembled, which 

 occupy little space, yet control the currents, 

 and deflect them upon the obstruction to be 

 removed. By it the prism of water passing 

 through a given section can be increased 

 indefinitely, while the aperture of discharge 

 may be diminished, thus producing any re- 

 quired velocity. Stress was laid upon the 

 importance of applying a method which 

 should be limited to the removal of so 

 much of the crest of the bar as would secure 

 the requisite channel, and no more. 



The Amazons Valley. Mr. James W. 

 Wells, in an address before the Royal Geo- 

 graphical Society, on "The Physical Geog- 

 raphy of Brazil," divides the rivers of that 

 country into three great systems : the basin 

 of the Amazons, including also the Tocan- 

 tins and Araguaya ; the basin of the Plate 

 River, and the many distinct and separate 

 rivers draining into the Atlantic. The Ama- 

 zons basin is divided into the bottle-shaped, 

 low-lying forest of the upper valley, 1,300 

 miles long by 800 miles broad, and its cir- 

 cumscribing elevated table-lands, which, near 

 Obidos and Santarcm, approach close to the 

 banks of the main river, and constitute the 

 neck of the bottle-shaped area. Throughout 

 the length of this river, east and west of 

 Obidos, the adjoining land is so low and flat 

 that we have in many cases rather a scries 

 of more or less parallel streams than one 

 great, clearly defined stream. It is possible 

 to go in a canoe up the whole of the valley 

 in these lateral channels, and also to pass 

 through the deep forest by natural canals, 

 from one tributary to another, without once 

 entering the main river. It is a singular 

 feature of the Amazons Valley, considering 



