130 



G. H. PARKER. 



Reasoning from this standpoint, the relative corrosive capacity in 

 seawater of the six metals used in these tests will be in the approxi- 

 mate order Zn, Al, Fe, Cu, Pb, and Sn. 



The poisonous effects of these metals on marine animals will 

 depend upon the intrinsic toxicity of their ions, relatively high for 

 all heavy metals, and the solubilities of their hydroxides and basic 

 carbonates in seawater. These solubilities in the case of Fe, Pb, 

 Sn, and Al are in amounts inappreciable ; in other words, these 

 metals in seawater are not surrounded by a layer of poisonous ions 

 and hence animals may grow upon them. In the case of Zn and 

 Cu, on the other hand, the corresponding compounds are appre- 

 ciably soluble in seawater and the poisons thus liberated prevent 

 the growth of animals upon these metals. Thus the presence or 

 absence of poisonous ions or compounds is what determines 

 whether a given metal will be covered with animal growth or not. 



TABLE II. 



COUPLES BETWEEN AL AND THE FlVE OTHER METALS, ALL OF WHICH STAND 

 BELOW AL IN THE ELECTROMOTIVE SERIES. 



The animals are indicated as in Table I. 



In the metallic couples the metal with the highest solution pres- 

 sure was Al. All the metals with which it was combined should, 

 therefore, render it active. So far as its solution pressure is con- 

 cerned, it is so near the next metal in the series, Zn, that it re- 

 mained, when combined with this metal in seawater, practically 

 uncorroded. In all other couples corrosion was evident and in- 

 creased in amount as the series was passed over from Fe to Cu 

 (Table II.). Since, however, the products of this corrosion are 

 practically insoluble, a general growth of animals in fair amounts 

 occurred on all the Al members of couples excepting one. In this 



