August 18, 1922] 



SCIENCE 



201 



then more recent globes of Bonne and Lelande 

 were sold in the one foot size at 100 livres 

 the pair; 10 inch at 18 livres; 8 inch at 10 

 livres ; 6 inch at 7. livi-es. The prices are more 

 than double those quoted by Stevenson (II, 

 136) from Moxon, a century earlier. 



Lalande cites among the enormous globes 

 one at Cambridge and one at Lyon by Picpus 

 de la Guillotiere. Lalande further states that 

 the finest of the large terrestrial globes is that 

 made in 1787 by D. Bergerin. The Cambridge 

 celestial globe was made, according to the 

 Encyclopedia Britannica (IX edition, Vol. X, 

 Globes, p. 683) about 1764 by Dr. Roger Long, 

 professor of astronomy and master of Pem- 

 broke College. This globe was 18 feet in diam- 

 eter, lined with tin. No one of these three 

 men is mentioned by Stevenson. 



Nowhere does the author touch upon early 

 appearances of globes in America, North or 

 South. One would expect to find the earliest 

 references in Mexico or Peru; certainly in 

 North America gloibes must have been import- 

 ed in the eighteenth century and possibly even 

 constructed here. In the geography by John 

 Payne, revised by James Hardie, that appeared 

 in New York in 1798, there is a figure of an 

 artificial sphere and pages xxxi-xxxviii are de- 

 voted to the use of the globe ; other references 

 could doubtless be found. 



These additions have been made to indicate 

 the wide appeal which globes have made in 

 the past as instruments of instruction. Steven- 

 son's work may well stimulate a revival of 

 interest in globes for instruction purposes. 



The two volumes constitute an enduring 

 monument of American scholarship. The 

 press work and the plates are up to the highest 

 standards of the finest presses of Europe. It 

 is to be hoped that students of astronomy and 

 geography in American colleges will make the 

 appearance of a second edition of more than 

 1,000 copies necessary. The author is to be 

 congratulated upon having added new laurels 

 to his crown in a field closely related to 

 cartography wherein the name of Edward 

 Luther Stevenson has so long stood first in 

 America and almost alone. 



L. C. Kaepinski 



University of Michigan 



SPECIAL ARTICLES 



THE EFFECT OF ABSORBED HYDROGEN ON 



THE THERMOELECTRIC PROPERTIES 



OF PALLADIUM 



It is well known that palladium will absorb 

 relatively large quantities of hycbogen under 

 the proper conditions. Palladium black ab- 

 sorbs the gas more readily and in larger quan- 

 tities than the solid metal but the latter will 

 contain several hundred times its own volume.' 

 The purpose of the work here described was to 

 determine the effect of the absorbed gas on the 

 thermoelectric properties of the metal. 



The palladium used was in the form of a 

 strip 0.01 X 0.125 X 10 cm. It was fii-st an- 

 nealed in a vacuum at a temperature of 1,000° 

 C, and then used in a thermo couple with a 

 strip of platinum as the other metal. The cold 

 junction was kept at 0° C. and the hot junction 

 oould be heated to various temperatures up to 

 300° C. This strip of platinum was used as a 

 reference metal throughout all of the deter- 

 minations. 



After the thei-moelectric power was obtained, 

 the palladium strip was heated to a tempera- 

 ture of about 700° C, in vacuo, and then 

 allowed to cool slowly in an atmosphere of 

 hydrogen. It is well known that palladium 

 will absorb hydrogen under these circumstances. 

 The thermoelectric power obtained with the gas 

 filled metal against platinum was less than with 

 gas free metal, amounting in one case, at 0° C, 

 for instance, to 73 per cent, of the gas free 

 value. The palladium was then heated in vacuo 

 to a temperature of about 700° C, to remove 

 the hydrogen, and another determination 

 showed the thermoelectric power to have re- 

 turned to its gas free value. This process was 

 repeated several times, the gas filled palladium 

 having its thermoelectric power against plati- 

 num lowered each time hydrogen was absorbed, 

 and restored again to its original value after 

 the hydrogen had been removed. 



A much greater decrease in the thermoelectric 

 power of palladium against platinum as a 

 reference metal was obtained when the palla- 

 dium was filled with hydrogen by the electro- 

 lytic method. The palladium strip was used as 

 the cathode in the electrolysis of water from a 

 very dilute solution of sulphuric acid. The 



