247 



consists of a bottle holding about two gallons with an opening 

 at the bottom provided with a valve. This is connected with a 

 bell-jar overhead holding about a liter and a half, and in the op- 

 eration one liter of water is allowed to pass very slowly from the 

 jar above into the lower one. This causes a displacement of air ; 

 in other words, one liter of air in the lower jar is forced out 

 through the tube shown at the left into the soil and the length 

 of time it takes this liter of air to pass through certain soils is 

 recorded. By using soils of different texture different values are 

 obtained. The soils in an experiment of this nature are usually 

 placed in metal cylinders of about 500 c.c. capacity, but lamp 

 chimneys provided with corks at the bottom may be substituted, 

 if necessary. 



Connecting the tube with the lower jar there is a water mano- 

 meter containing an inch or two of water in each arm, which 

 determines the pressure of the air due to resistance to pressure 

 through the soil, and when one liter of air is passed from the 

 upper bell-jar into the jar below and the water columns remain 

 precisely on a level, exactly one liter of air has been passed 

 through the soil. The contrivances often used for this purpose 

 are so clumsy and constructed on such poor mechanical prin- 

 ciples that it is impossible to get the same results from the same 

 soil twice, but by the use of the device described above, very 

 reliable results are obtainable. 



Massachusetts Agricultural College. 



Mr. J. A. Udden, of Rock Island, Illinois, reports in Science 

 for July 31, a fossil cycad found in the Upper Cretaceous of Texas 

 Eight fragments of what was presumably the same silicified trunk 

 were found, three of these matched by their fractures and showed 

 a stem about ten inches wide, hollow, and considerably flattened. 



The Century for September has an illustrated article on the 

 Future Wheat Supply of the United States, written by Edward 

 C. Parker, Assistant at the Agricultural Experiment Station of 

 the University of Minnesota. The methods of the wheat breeder 

 are clearly described, and besides the economic information indi- 



