374 W. M. Fontame— Conglomerate Series of West Vvrginia. 



particles of the epidermal shield, there was observed a marvelous 

 interlacing of the various forms, showing that they were con- 

 veyed by the sap ceils directly to the section of the plant 

 where they were destined to complete its structure. I have 

 examined several specimens of straw, taken at random in the 

 market; the silica in each specimen consisted of plates, very 

 thin, and truncated at the comers. 



The result of these investigations shows the necessity of fine- 

 ly divided silica in the soil, so minute as to be capable of pass- 

 ing with facility through the sap cells ; secondly, that simple or 

 compound silicates are useless as fertilizing agents, either nat- 

 ural, or artificially prepared. We have no valid reason for 

 forming any theory that vegetation can, through any known 

 chemical law, separate the elements or their compounds from 

 combinations so positive in their character. 



In this case we have a practical result capable of being veri- 

 fied at any stage of growth of a plant, produced by the appli- 

 cation of silica to the soil in the form of certain well defined 

 microscopic organisms ; for, finding these in the ash to the ex- 

 clusion of other particles of silica, they seem to be more accept- 

 able for the plant structure. Free silica is hence the only 

 condition in which it can enter the plant. 



I look upon this discovery as leading agricultural investiga- 

 tions in a new direction, and it must eventually change many of 

 the views expressed and accepted by scientists. 



Every precaution was used in having all the material thor- 

 oughly cleansed, with a view both for accuracy and for remov- 

 ing suspicions that these microscopic forms were the result of 

 dust showers. 



[Concluded from page 284.] 



Dr. Stevenson, in his " Notes on the Geologv of West Vir- 

 ginia," (read before the Am. Phil. Soc, Feb. 5, 1875,) speaks as 

 follows of the " Great Conglomerate" of Kandolph county. " This 

 rock forms the crest of Rich Mountain for nearly sixteen miles 

 within the region examined. For the most part it is a coarse 

 sandstone, loaded with pebbles from one third of an inch to two 

 inches in diameter. Along the Staunton Pike it shows some 

 layers of slightly micaceous and very compact sandstone near 

 the bottom. Here it is greatly increased in thickness ; near the 

 northern line of the state it is barely Zk>Q feet thick, but in Ran- 

 dolph county it is not less than 600." He further says, " On the 



