446 



NATURE 



[December 19, 1912 



up a mass of cold rock as against one that has long 

 felt the influence of the sun ; hence unusual precipita- 

 tion follows on the new land-surface, and an Ice age 

 sets in. Stanislas Meunier (" Les Theories 

 Glaciaires," Revue des Idies, igio, p. 207) affirms, 

 as usuaj, that no general and contemporaneous re- 

 frigeration has been proved ; but he also asserts that 

 the scratches of stones in boulder-clay are produced 

 by the infiltration of rain and consequent settling of 

 the mass. M. Yokoyama ("Climatic Changes in 

 Japan since the Pliocene Epoch," Journ. Coll. Sci., 

 Tokyo, vol. xxxii., 191 1, part v.) cannot accept the 

 evolution of carbon dioxide as a cause of warmer 

 cliipates, since in Japan the output must have been 

 considerable during glacial times. He prefers, from 

 local palsontological evidence, to account for the 

 difficulties by a shifting of the poles. This is, of 

 course, seriously opposed by the evidence of contem- 

 poraneous world-wide refrigeration. R. Speight 

 ("The Post-glacial Climate of Canterbury," Trans. 

 New Zealand Inst., ,'ol. xliii., 191 1, p. 408) finds no 

 local cause in New Zealand to account for the suc- 

 cession of climates that he records, a moist climate 

 following the glacial, and modified steppe conditions 

 preceding those of the present day. The author 

 points out that the sequence is so similar to that in 

 Europe as to suggest some cause that affected the 

 whole earth, though changes in the grouping of 

 land and water in the southern hemisphere might 

 account for the former conditions in New Zealand. 



G. A. J. C. 



T' 



THE WORK OF THE PHYSIKALISCH- 

 TECHNISCHE REICHSANSTALT, 

 CHARLOTTENBURG, IN 191 1. 

 'HE following notes describe some of the more 

 important researches, &c., undertaken at the 

 above institution during 191 1. They are compiled 

 from the annual report of the Reichsanstalt, appear- 

 ing in Zeitsckrift fiir Iiistrumentcnkunde, April, May, 

 and June, 1912. 



The comparison of platinum resistance thermometers 

 with various gas thermometers has been completed 

 between 0° and 450° C. It was found that the 

 hydrogen thermometer and the helium thermometer 

 of constant volume with an initial pressure of 620 mm. 

 mercury indicated about o'i° higher at 450^^ C. than 

 the nitrogen thermometer under the same conditions. 

 With the accuracy attained, the hydrogen scale may 

 be at once identified with the ideal scale, since, ac- 

 cording to Berthelot, these only differ by about o'oi" 

 at 450 C. in the present case. The data for the fol- 

 lowing fixed points, which were determined afresh, 

 refer to the ideal gas scale : — 



Freezing points Boiling points 



Tin 23i'8,° Naphthalin ... 2i7'9° 



Cadmium ... szo'g^'^ Benzophenon... 30S'8,° 

 Zinc 4i9'4o" Sulphur ... 444'5i° 



In connection with an investigation of the mean 

 specific heat of gases at high pressures, the specific 

 heat of air between 20° and 100° C. at i and 11 atmo- 

 spheres was measured with a new calorimeter. It 

 was found that when the pressure was increased from 

 I to II atmospheres the specific heat increased by 

 about 2' I per cent. This result must not, however, 

 be considered as final at present. 



The investigation into the specific heat at constant 

 pressure of air by the Callendar and Barnes con- 

 tinuous-flow method was concluded. The values 

 found for the specific heat at constant pressure of 

 carbonic-acid-free air under atmospheric pressure are 



NO. 2251, VOL. 90] 



given below. The method gives the results direct in 

 electrical measure (watt-seconds) ; and the values con- 

 verted into heat units (C;il.i:,,: are also given. 



- 78 



-183 



The experiments were extended to carbonic acid gas, 

 o.\ygen, and nitrogen. For the pure, dry gases, at 

 atmospheric pressure and 20° C, the following results 

 were found in electrical and in heat units respec- 

 tively : — 



Carbonic acid gas 



Oxj gen 



Nitrogen 



<',j = o'846 and o'202 respectively. 

 (-,, = o'9i7 „ o'2i9 „ 



t>=ro4i „ o'249 ,, 



For carbonic acid gas at —78° C. and atmospiieric 

 pressure the respective results were, c,, =o'765 and 

 o'i83. The decrease in specific heat of CO^ between 

 + 20'^ C. and —78° C. is, when calculated per degree, 

 only slightly less than that between + 100° C. and 

 + 20° C. determined by Swan. 



Specific Heat of Water between o"^ and 100° C. — A 

 precise determination of the calorie in electrical units 

 on a trustworthy basis appears very desirable. The 

 bases of the measurement, vjz. the unit of resist- 

 ance, the e.m.f. of the standard cell and the tem- 

 perature scale, have now been fixed internationally to 

 such a degree of certainty as to appear to render 

 possible a determination of the calorie in international 

 watt-seconds to within i part in 10,000. This re- 

 search was commenced at room temperature, and a 

 description of the various apparatus and of the experi- 

 mental arrangements is given in the report. No 

 results are, however, recorded. 



Weston Normal Cells. — A number of these were 

 constructed, using new mercurous sulphate prepara- 

 tions, with the view of seeing whether all freshly pre- 

 cipitated samples yielded the same e.m.f. as the older 

 preparations, and for the purpose of discovering 

 whether the method of washing the precipitated mer- 

 curous sulphate had any influence on the e.m.f. of 

 the cell. The results show that the method of wash- 

 ing has no appreciable influence on the e.m.f. Other 

 extensive investigations were undertaken on Weston 

 cells, and the general results arrived at indicate that 

 both the reproducibility and constancy of the cell can 

 be guaranteed internationally to within a few parts 

 in 100,000. 



In connection with some experiments un resistance 

 thermometers, it was found that the differences shown 

 between fused silica platinum resistance thermometers 

 and the ordinary type may bo ascribed to a reaction of 

 the quartz glass on the platinum — probably of a 

 chemical nature. Experiments were also made with 

 the view of comparing the behaviour of the quartz 

 glass resistance thermometer at the highest tempera- 

 tures at which it can be used with the ordinary resist- 

 ance thermometer. Full details of these experiments 

 are given. 



Electrolysis of Glass. — ^The investigation of the 

 badly conducting layers discovered by VVarburg in the 

 electrolysis at 3oo°-35o° C. gave the following 

 result : — 



Platinum or graphite anodes are not soluble in 

 glass. On electrolysis, a layer of high resistance 

 occurs at these anodes, sodium migrating from the 

 glass to the kathode and oxygen to the anode. With 

 mercury as anode, quantitative migration takes place. 

 The metals lead, bismuth, antimony, tin, iron, and 

 copper, when oxide-free, appear to migrate quantifa- 



