Tebeuaet 2, 1917] 



SCIENCE 



121 



of Cuba than from South America by way of 

 the Lesser Antilles; certainly not from North 

 America. But it seems doubtful whether any 

 former continental connection is indicated, the 

 mammalian fauna, like that of Cuba, etc., be- 

 ing limited to a few groups which can be ac- 

 counted for in other ways. 



Mr. E. L. Troxell-^ describes the skeleton of 

 a Pliocene horse which is in many respects in- 

 termediate between the three-toed horses of 

 the Miocene and the true Equus of the Pleis- 

 tocene. It is referred to the genus Pliohippus, 

 but is much more complete and more truly in- 

 termediate in character than the type species 

 described many years ago by Marsh. A second 

 and more complete skeleton has recently been 

 discovered in western ^Nebraska; both are in 

 the American Museum in 'New York. [W. 

 D. M.] C. E. Eastman, 



W. K. Gregory, 

 W. D. Matthew 



SPECIAL ARTICLES 



THE REFLECTION OF T-EAYS BY CRYSTALS' 



Rutherford and Andrade- have shown that 

 when 7-rays fall on the faces of crystals at cer- 

 tain angles regular reflection takes place as in 

 the experiments of Bragg^ with X-rays. This 

 should show itself by an increase of absorp- 

 tion of the 7-rays, and in the experiments to be 

 described evidence has been obtained of this 

 character. 



A fine pencil of 7-rays passed through a 

 vessel containing a crystalline substance into 

 an ionization chamber where the ionization 

 was measured. The crystalline structure of 



simmer. Jour. ScL, Vol. 42, pp. 335-348, 7 text 

 figs. 



1 This article was written in April, 1914, and de- 

 scribes some experiments performed in Professor 

 Sir Ernest Eutherford's laboratory at the Uni- 

 versity of Manchester. At that time Eutherford 

 and Andrade were working on the same problem 

 by the more direct method. While the results re- 

 corded in this paper have apparently little quanti- 

 tative value, the general method of attack may be 

 of sufScient interest to justify their publication. 



2 Eutherford and Andrade, Phil. Mag., May, 

 1914, p. 854. 



3 Bragg, Phil. Mag., May, 1914, p. 881. 



the absorber was then destroyed either by 

 powdering, melting or by dissolving in water, 

 and any change in the ionization current was 

 measured by a balance method. The change 

 in the ionization gives a measure of the radia- 

 tion which is reflected from the crystals at 

 such an angle with the direction of the beam 

 as not to enter the ionization chamber. The 

 experimental arrangement is shown in Pig. 1. 



The small thin glass crystallizing dish D, 

 containing the crystals under investigation was 

 placed over a hole in the lead block L so as to 

 rest either directly on the lead block or on an 

 adjustable iron-gauze shelf above it. The 

 7-rays from the source S passed through the 

 crystals and hole, which was 1.2 cm. in diam- 

 eter, and through a very thin sheet of alumi- 

 num foil into the ionization chamber A. The 

 balance chamber B also received 7-rays from 

 the source through a thick adjustable lead 

 slit B. Electrodes passing into A and B 

 through earthed guard rings were connected 

 to a Wilson-Kaye electroscope E. The cham- 

 bers A and B were hollow brass cylinders 15 

 cm. long and 8 cm. in diameter. They were 

 insulated and connected to — 200 volts and 

 -\- 200 volts, respectively. By means of the 

 key K the gold leaf could be earthed or joined 

 to a divided megohm in series with a storage 

 battery for the purpose of measuring the sen- 

 sibility of the leaf. The leads to the electro- 

 scope from the chambers A and B were com- 

 pletely shielded by brass tubing and lead foil 

 earth connected, so that electrostatic effects 

 were eliminated. The balance chamber B was 

 surrounded by a lead sheet 3 mm. thick to 

 prevent any soft scattered radiation from 

 entering it, and all connections to the electro- 

 scope were shielded as much as possible from 

 direct radiation by thick blocks of lead. The 

 lead block L was 7.5 cm. thick, and for the 

 position of the source used in most of the ex- 

 periments about twenty-five times as much 

 ionization was produced by the rays passing 

 through the hole as through the rest of the 

 block. 



Owing to their short wave-lengths the angles 

 of reflection for 7-rays are probably small. It 

 is, therefore, necessary to use a small cone of 



