liv INTRODUCTION. [CH. 



laboratory a piece of earthenware, which was used merely 

 to hold a small voltaic battery and catch the chance 

 spillings of sulphate of zinc from the cells, which is 

 here and there eaten out in deep pits, and in other places 

 pockmarked with small spots. If this dish had been 

 exposed to alternations of sea- water and air, a similar 

 effect would have taken place ; and yet there was no per- 

 ceptible want of homogeneity in the dish at first. 



" The tree-like corrosions between the object-lenses of 

 old telescopes are probably due to the same causes, or one 

 of them. The old experiments of the zinc and silver 

 tree are instances of the same sort of action. Wet a 

 glass plate with nitrate of silver and hold the point of a 

 pin in it, the acid leaves the silver for the copper ; but 

 the silver is not deposited in a uniform circle, but in a 

 beautiful arborescent form. Old wine-bottles are fre- 

 quently found corroded, some in spots^ others in tree- 

 like figures. 



" I believe the above will help to solve the problem 

 you are investigating ; at all events, I can offer no better 

 solution. ff Ever yours, 



" W. R. Grove.^^ 



I may add that limestone rocks are fretted in the 

 same way as the shells of Purpura lapillus and a stunted 

 variety of Mytilus edulis which are found on these rocks, 

 all having the same calcareous basis. The erosion of 

 bivalves is greater at their beaks, where the connexion 

 between the animal and the shell is weaker than in other 

 parts. This is also the case with the points of univalves. 

 The action appears to be prevented by the epidermis. 



Operculum. — The horny or shelly plate ("pot-lid") 

 by which many univalve Mollusca close the mouths of 

 their shells is attached by a strong muscle to the back 

 or upper surface of the foot ; and it serves to protect the 



