lii INTRODUCTION. [CH. 



to his examination were those of Littorina litorea, or the 

 common periwinkle. 



" March 26, 1862. 



^' Dear Jeffreys, — As you wish me to put on paper 

 the suggested explanation of your difficulty, viz. why 

 it is that substances apparently homogeneous are cor- 

 roded in patches or irregular hollows, I endeavour to do 

 so, premising that it is only theoretical and may be 

 found not to accord with all vour facts. 



" If a plate of pure zinc be immersed in dilute sul- 

 phuric acid, little or no chemical action takes place ; 

 but keep in contact with the zinc another metal, say an 

 iron nail, and the zinc is rapidly oxidated and dissolved, 

 hydrogen being evolved from the surface of the iron. The 

 action is most rapid at the points most proximate to the 

 iron ; so that, if the nail be laid upon the zinc, both being 

 immersed in the liquid, the iron would seem to eat its 

 way into the zinc. If a few iron filings be placed upon 

 the zinc, the action will be similar, but Avill begin more 

 slowly and increase as the points of iron are exposed, 

 these latter Ijeing protected to a great extent by their 

 being negative and coated T\ith hydrogen. If an in- 

 finitesimal quantity of copper be dissohed in the acid, it 

 will be precipitated in a metallic state on the negative 

 points and make these permanent centres of action. So, 

 if no iron be made to touch the zinc, after a time some 

 want of absolute homogeneity is sure to determine a 

 chemical action ; and as any trace of metal existing in 

 the solution will by this chemical (or, rather, electrolytic) 

 action be deposited on the negative points, or those 

 least attacked, the action will continually increase, and, 

 instead of being uniform, mil be in patches around the 

 negative centres. Thus a piece of common zinc of com- 

 merce which contains small portions of iron and tin will 



