OF THE FREE DIATOMACEAE. 21 



separated themselves, and sunk to the bottom. This will 

 take place in about half a minute. The fluid is next 

 carefully decanted into a second glass, and the amount of 

 water lost made up, the same process being gone through 

 five or six times. The sediment in the second glass now 

 contains all the Diatomaceae and sandy particles, which 

 were too light to subside during the first half minute. To 

 separate these still further, this sediment is treated exactly 

 as was the original gathering, excepting that a longer time 

 is given it (say two minutes and a half), between each 

 decanting. In the same way the contents of the third 

 glass are sorted, only that a still longer period, not less 

 than five minutes, is allowed. Every glass now contains 

 samples of Diatomaceae, mingled with earthy ingredients of 

 varying weights, the last glass having of course the 

 lightest. To remove the foreign bodies altogether, each 

 mass of sediment is placed in turn in a short wide glass, 

 and about an inch height of water is poured over it, where 

 it is left for a minute or two. As soon as the whole of the 

 contents may be supposed to have settled, a rotatory 

 motion is given to the water, by moving the vessel with 

 the hand in a circular direction. The Diatomaceae, con- 

 sisting mostly of thin plates or valves, are raised by the 

 agitation of the water to the surface, while, on the other 

 hand, the sandy particles befng more or less round, are 

 rolled upon each other at the bottom, and gradually collect 

 in the centre of the vessel. The fluid is now quickly, but 

 with great care, decanted into a second glass, where the 

 Diatomaceae are allowed to subside. What remains in the 

 first glass may be again and again submitted to the same 

 process, until it is tolerably certain that not a single 

 Diatom has been overlooked. 



In order to effect a separation of the different species, 

 by taking advantage of their varying specific gravities, 

 the following method, introduced by Munro, is excellent, 

 and has the merit of great simplicity. A glass tube, 

 three to four feet in length and half an inch wide, is 



