MEMORANDA. 159 



call the last remaintler, or what remains in A after it has 

 stood its half-hour, " Not in five minutes." Thus we have 

 four different glasses, containing diatomes and clay mixed, of 

 four different densities : thus to ^ ; ^ to 2i ; 2i^ to 5 ; not 

 in 5. There is now a method of concentrating the coarsest of 

 these sediments, namely, the to i, the ^ to 2+, and some- 

 times the 2^ to 5. It consists in taking the beaker containing 

 the sediment and pouring about an inch of water on it. Let 

 it settle about 5 minutes and then place the glass on a table, 

 and impart a whirling motion to the whole by moving it 

 round and round, when the greatest portion of the diatomes 

 will rise up in a sort of eddy, while the particles of mud or 

 sand will remain at the bottom, even though they are of the 

 same specific gravity as the diatomes, and have fallen through 

 the same distance of water in the same time. This is because 

 the diatomes are mostly Jlat find thin, while the particles of 

 sand and mud are round ; in the same way, if we take a round 

 pebble and an oyster shell both of the same weight and throw 

 both horizontally into the water, the pebble will reach the 

 bottom sooner than the oyster shell. So when the whirling 

 motion is imparted to the glass, the thin flat shells of the 

 diatomes will rise up in a cloud, while the round particles of 

 mud and sand will remain behind ; when the cloud rises up, 

 pour it off quickly and dexterously into another glass, and, if 

 necessary, repeat the process, and a little practice will enable 

 the operator to separate all the diatomes most effectually. I 

 have said before that this process will only apply to the 

 to ^, ^ to 2t, and sometimes the 2i to 5 sediment, but not at 

 any finer one ; practice may soon teach this. The " not in 5" 

 cannot be concentrated, it is too fine, and the whole rises 

 together on imparting the whirling motion to it. 



It is not necessary to abide invariably by the divisions of 

 time which I have given here. 



These must be varied, of course, according to the nature of 

 the clay to be examined. For instance, in a clay I have re- 

 cently tried from 34 feet below the bed of the river at Cardifl", 

 nearly the whole of what was left after the to ^ fell in the 

 T to 2^. I, therefore, divided it thus : to i, ^ to li, and 

 1^ to 2^ ; a little practice will soon teach this. 



The advantages of the plan are, 1 think, obvious. In the 

 first or coarsest sediments we get all the larger and finer 

 Diatomes by themselves, unmixed with and consequently un- 

 obscuied by the innumerable smaller ones, and the fine 

 particles of mud and sand ; while if any of them, such as the 

 Enpodisci or Campj/lodisci are rare, they are sure to be found 

 in either the first or second division of densities, and by their 



