616 



SCIENCE 



[N. S. Vol. XXVIII. No. 722 



sand screen with eighit meshes to the inch 

 is placed inside of the inner steel cylinder. 

 The screen or wire cylinder should fit into 

 position perfectly and there should be no open 

 space between this cylinder and the inner steel 

 cylinder. As the outer cylinder bores into the 

 soil and separates a core of soil from the soil 

 mass, the inner steel cylinder, carrying the 

 wire cylinder is carried downward at a rate 

 uniform with that of the outer cylinder and 

 the core of soil is pushed with but little fric- 

 tion and in an unbroken condition into the 

 wire cylinder. When a sample of soil has 

 been secured to the desired depth, the sampler 

 is withdrawn and the wire cylinder which con- 

 tains the core of soil is removed from the 

 machine. When the soil sampler is in opera- 

 tion, it is held rigidly in position by a wooden 

 frame which is supported on four legs. 



In conclusion it may be said that the ad- 

 vantages which the writer thinks should com- 

 mend this new apparatus for taking soil 

 samples and particularly those which are used 

 for the determination of the physical char- 

 acteristics of the soil, are the rapidity with 

 which samples can be secured, and the un- 

 changed physical condition of the core of soil. 



The claim is not made for this method that 

 the samples duplicate closely when tests are 

 made regarding the physical properties of a 

 soil type. However, it is the opinion of the 

 writer that the variations are due wholly to 

 factors other than those connected with the 

 operation of securing the samples of soil, and 

 it is not probable that these factors can be 

 eliminated. 



W. H. Stevenson 



Iowa State College, 

 Ames, Iowa 



NOTE ON the crystal FORM OP BBNITOITE 



Of the thirty-two possible crystal classes 

 deduced mathematically from the empirical 

 law of rational indices by Hessel in 1832, 

 three have no known representative up to the 

 present time. They are the tetragonal bi- 

 sphenoidal, trigonal bipyramidal and ditrig- 

 onal bipyramidal classes. The writer be- 

 lieves that the last-mentioned class, the ditri- 

 gonal bipyramidal, has a representative in the 



new gem mineral, benitoite (BaTiSjO,) re- 

 cently described by Louderback.' 



Several crystals of this interesting mineral 

 obtained through E. M. Wilke were examined 

 and measured, with the following results. The 

 dominant form is a trigonal bipyramid, which 

 determines the habit. If this is taken as the 

 positive unit form, 1011, the other forms 

 (taking the axes of reference diagonal to the 

 planes of symmetry as in tourmaline) are: 

 0111 and 0112 trigonal bipyramids; 1010 and 

 0110, trigonal prisms ; and 0001, pinacoid. Of 

 these 0111 is small, 0112, a narrow form 

 truncating the polar edges of 1011 and only 

 found on one or two crystals. Of the two 

 prisms 1010 is invariably the more promi- 

 nent, but 0110 measures a little more in the 

 direction of the c-axis. The pinacoid 0001 is 

 a small triangular face and on one crystal 

 there were triangular markings parallel to its 



Although the general form, hMl, ditrig- 

 onal bipyramid, is absent, it is pretty certain 

 that the crystals belong to the class mentioned 

 as there is a horizontal plane of symmetry in 

 addition to three vertical planes of symmetry 

 and three axes of two-fold symmetry as well 

 as a single axis of three-fold symmetry. 



Another possibility is that the crystals may 

 belong to the trigonal bipyramidal class in 

 which case the dominant form would be an 

 hTcll face, but limit forms are much more com- 

 mon among crystals than general forms. It 

 may also be urged that the crystals may be 

 supplementary twins of the ditrigonal scalen- 

 ohedral or of the ditrigonal pyramidal class, 

 but as the prism faces show no grooves, 

 nicks, striations or seam through the center, 

 it seems reasonable to regard them as simple 

 crystals. 



Sufficient angles were measured to establish 

 the forms as given above. The average of 

 ten values for the angle (0001 A 1011) varying 

 from 40° 0' to 40° 22', gave 40° 10' as com- 

 pared with Louderback's value of 40° 14'. 



Austin F. Eogers 



Stanford Univebsity, Cal., 

 September 19, 1908 



^Bull. Dept. Geol. Vniv. Cal., Vol. 5, pp. 149- 

 153, 1907. 



