Physics. 



329 



(1847) ; Pierre (1845-52) ; Hagen (1855) ; and Matthiesen 

 (1866). These six series of determinations were also combined 

 by the method of Schiapparelli and the resultant curve Ros- 

 setti adopted as the most accurate attainable from existing 

 data. From this curve, in which 1 mm. of ordinate equalled 

 of the volume of water at 0° C, he obtained the fol- 



100000 



lowing formula which represents the curve quite closely. 



Y t = 1 + a(t - 4)- — b(t - 4) 2 ' 5 + c(t - 4y 

 when 

 a = -00000837991 ; b = -000000378702 ; o = -0000000224329. 



From Rossetti's table of the relations of the volume, density 

 and temperature of water we extract the following data which 

 are all that will be required in their application to our method 

 of determining the coefficient of cubical expansion. 



Temp. C°. 



Volume. 



Temp. 



C°. Volume. 



Temp. C°. 



Volume. 







1-000000 



14 



1-000572 



28 



1003553 



1 



•999943 



15 



1-000712 



29 



1-003835 



2 



•999902 



16 



1-000870 



30 



1-004123 



3 



•999880 



17 



1-00] 031 



31 



1-00442 



4 



•999871 



18 



1-001219 



32 



1-00473 



5 



•999881 



19 



1-001413 



33 



1-00505 



6 



•999901 



20 



1-001614 



34 



1-00538 



7 



•999938 



21 



1-001828 



35 



1-00572 



8 



•999985 



22 



1-002049 



36 



1-00608 



9 



1-000047 



23 



1-002276 



37 



1-00645 



10 



1000124 



24 



1-002511 



38 



1-00682 



11 



1-000216 



25 



1-002759 



39 



1-00719 



12 



1-000322 



26 



1-003014 



40 



1-00757 



13 



1-000441 



27 



1-003278 







Stevens Institute of Techi 



iiology, Hoboken, N. J. 









SCIENTIFIC 



INTELLIGENCE. 









I. 



Physics. 







1. Steam Calorimeter. — K. Wirtz has modified the steam 

 calorimeter of Bunsen and Jolly and has shown how the method 

 of these investigators can be employed to measure the latent heat 

 of vapor of substances with low boiling points. A weighed 

 amount, G, of the substance is vaporized in the steam of the 

 calorimeter. Heat is withdrawn from the steam and a portion of 

 the latter is condensed. Calling to the weight of the condensed 

 steam, X the latent heat of steam, the latent heat, Q, of the 

 vapor of the substance between the temperature before it is put 



into the calorimeter and its boiling point will be Qz=-^-. The 



author discusses the sources of error. The results obtained by 

 him are in general smaller than those obtained by Regnault, 

 Person and Andrews. This failure of agreement is attributed 

 by the author to the use of impure substances by the previous 



