■■To have gold in a shell, it is not necessary it should be an oxide. It is only 

 necessary it should have been received into the circulation of the animal, in 

 solution as chloride, or some other possible soluble form that chemistry has not 

 brought to light; and -when once in the circulation it may be eliminated by be- 

 ing deprived of its solving principle and excreted or secreted with the other 

 solid matter that enters into the formation of the shell. Thestannateof gold, or 

 purple of Cassius, may be wholly deprived of the tin associated with it. yet re- 

 tain it; purple color, and its condition of atomic division, if so you are pleased 

 to call it. But I only offer this as suggestive of something for those interested 

 :low further. I am not enough of a chemist to develop any facts out of a 

 suspicion of this kind." 



Mr. Lea remarked, after reading the above extracts, that the purple, pink and 

 salmon color of many of our American Unionidce had had his attention from the 

 period of his first studying this beautiful and interesting family, more than thirty 

 •car; since. Without having experimented himself upon them, he was aware 

 that no chemist had been able to detect the presence of a metal or other 

 elementary body. He therefore thought it likely to be caused by the presence 

 of some organic body which had not yet been detected; such is supposed by 

 imiststobe the case with the colored fluates of lime, colored quartz, &c. 

 What Dr. Lewis states as regards the colors being more frequent and more in- 

 tense in the waters of Michigan and in the streams leading into the northern 

 great lakes from the southern side, is very true. The TJnio rectus is usually 

 white in the Ohio, though sometimes tinted with purple and salmon color, while 

 in the more northern waters it is usually of a fine rich purple or salmon. Two 

 specimens from the upper Mississippi, brought by Dr. Cooper, were exhibited 

 Mr. Lea, which were of exquisite purple and salmon. The Unio ligamentimis 

 has probably never been found pink or purple in the Ohio, while at Grand 

 Rapid3, Michigan, those with a fine pink and salmon color are very common. 

 The iLiryiirittino margarUifera of Columbia river and its tributaries has a fine 

 purple nacre in almost all the specimens, rarely white, while those in the rivers 

 •if Pennsylvania. Connecticut and Massachusetts are almost universally white, 

 as those from the northern part of Europe are also. 



Dr. Draper had informed Mr. Lea that he had calcined some of these purple 

 shells, but that they had burned white and he had not detected any metallic sub- 

 tance in their composition. The subject was certainly one well worth the pursuit, 

 as no doubt could remain that the color was derived from some foreign sub- 

 stance entering into the composition of some individuals, while others were 

 free from it. It was not an uncommon case to find the dorsal portion of 

 the nacre to be pink or purple while the other portions were white, and this was 



