Febeuakt 28, 1908] 



SCIENCE 



353 



of absorption for air is very small, the pene- 

 trating radiation would be very considerably 

 increased. The amount of ionization in the 

 open air is in agreement with this view. 



Curve 2 is very similar to the curves repre- 

 senting the hourly rate of change of tempera- 

 ture during the day. 



The third method is the determination of 

 the rate of leak below the ground. If the 

 penetrating radiation is due to the y radia- 

 tion from radium in the gTound, the ioniza- 

 tion caused by it should be twice as much in 

 a well or cave in the ground. During the 

 summer the writer tested this by placing the 

 electroscope in a cave. The rate of leak was 

 found to be somewhat less than that occurring 

 during the night on the surface, and was 

 found to be about the same as the leak when 

 the electroscope was surrounded with four feet 

 of water in a cistern. Moreover, the rate of 

 leak was the same during the day as during 

 the night. Similar results have been found 

 by Elster and Geitel. 



The conclusion is that a large part of the 

 ionization in closed vessels due to the ex- 

 ternal penetrating rediation is subject to 

 large variations, and probably consists of y 

 rays from radioactive products in the air 

 rather than in the ground. These products 

 vary very much in quantity according to 

 atmospheric conditions. A very convenient 

 method of measuring these products is by 

 measuring the amount of the ionization due to 

 the penetrating radiation. 



The writer wishes to express his sincere ap- 



preciation of the constant spirit of interest 

 shown by Professor Ames. 



W. W. Strong 

 Johns Hopkins Univeesitt 



melting-point curves op binary mixtures of 

 the three nitranilines 

 Some interesting results have developed in 

 the course of an investigatioii of the subject 

 mentioned above. We prepared a series of 

 mixtures of ortho- and metanitraniline, the 

 difference in composition between each con- 

 secutive member being 2 per cent., and plotted 

 the melting points of these mixtures against 

 their composition. The curve produced was 

 regular, it fell from 71°, the m. p. of the pure 

 ortho compound, to 54°, the eutectic point, 

 then rose steadily to 114°, the m. p. of pure 

 metanitraniline. It is possible, therefore, to 

 employ the curves to determine the composi- 

 tion of mixtures of the two isomers. For this 

 purpose the melting point of the substance 

 under examination is determined; this gives, 

 in general, a temperature corresponding to a 

 position on each limb of the curve and there- 

 fore to two possible mixtures. Another por- 

 tion of the substance is now gi'ound up with 

 an approximately equal weight of either of the 

 pure isomers — ^say the meta- and the m. p. 

 of this mixture is determined. If it is loiver 

 than that of the first one the composition of 

 the original material is represented by the 

 point on the ortho- limb of the curve, whereas, 

 if the second m. p. is higher than that of the 

 first, the composition is read oif from the 



