THE MICROSCOPE. 3 
moderate shaking will settle clear in one minute, after a vigorous > 
shaking will not clear in ten: therefore shake. Allow to settle ten 
minutes, draw off water with siphon and renew. Repeat this until 
perfectly clear. 
Next attack that most terrible bugbear, the fine sand. Take a 
shallow glass dish with very slightly concave bottom (a photo- 
grapher’s “bender” is most suitable), and place in it a quantity of 
the material not sufficiently great to heap up much. Now, the friec- 
tion between water and glass being infinite, it may be conceived that 
there is a layer of water next to the glass which does not move when 
the glass is tipped, and the glass must be so manipulated that the 
fine sand will be retained in this layer, and the diatoms swept over 
it, and gathered at a point where they may be drawn off with a 
pipette. Only experience will show how this is best done, but it is 
possible to effect a perfect separation by rocking and tipping and 
shaking gently from side to side. As the diatoms are separated from 
the sand, draw them off with a pipette, add more water and continue 
until no diatoms are left. Throw away the sand, return the diatoms 
to the glass and repeat the whole process until the sand is all removed. 
There may still remain with the diatoms a little of the fine amor- 
phous matter not removed in the ten-minute settlings, and some par- 
ticles of dust which have entered during the long exposure to the 
atmosphere. To dispose of these, again place the material in the 
wide-mouthed vial with filtered water, add a few drops of aqua 
ammonia, and shake. Now, if it is desired to separate coarse from 
fine forms, or to eliminate broken frustules, it can be done to a certain 
extent by microscopic examination of the settlings, and timing them 
accordingly. In general, allow to settle until all forms desired in a 
given settling are precipitated, draw off the water into a larger 
vessel, fill up the vial, shake and settle the same length of time as 
before, and continue until everything which will not settle in that 
time is washed out. That material will then be finished. Then take 
the residue, shake and settle enough longer to deposit the next 
smaller forms desired. Treat these as before, until all smaller than 
these are washed out, and so on until nothing is left, which is desir- 
able. Remember that forms which settle, say in one minute, will 
carry down with them many smaller forms which, after repeated 
washings of the original one-minute deposit, will be released. It is 
only by tireless assiduity in shaking and settling that anything like a 
complete separation can be obtained. 
If it is not desired to separate the different forms, but only to 
remove any fine particles which may remain, simply shake the vial, 
