The perfectly dried matter is best treated in exactly the same way as a residuum in water analysis.  It is a common thing to ignite the residuum, and to put the loss down, if any, to water.  This ought not to satisfy an accurate observer, since organic matter, carbonates—­especially in presence of silica—­will easily add to the loss.  The best plan is to heat a small portion very cautiously, and note if any smell or alteration in color, due to carbon, etc., is perceptible, and to proceed accordingly.

I have seen some very satisfactory analyses made on board ship by a skillful use of the blowpipe, where liquid reagents would be very inconvenient to employ.

It will be necessary to say a few words as to the way in which soundings are made at sea.  When the bottom consists of sand, mud, or other loose matter, it is easy enough to bring specimens to the surface, and, of course, we know in such a case that the bottom has been reached, but, in the event of the bottom being hard and rocky, it is not easy to say that our sounding has been successful:  and here we meet with a difficulty which unfortunately is most unsatisfactorily provided for.

The lead is “cast,” as the saying goes, “armed” for this emergency.  An iron sinker is made with a hollow recess in the bottom; this is filled in with tallow, and on striking the bottom any loose matter may adhere by being pressed into the tallow.  If the bottom is rocky or hard we get simply an imprint in the arming, and when such a result is obtained the usual construction is that “the bottom is rocky” or hard.

Now, this seems to me a point on which chemistry may give some very valuable help, for I am convinced that no sounding should be accepted unless evidence of the bottom itself is obtained.  A few considerations will show that when we are working in very deep water, where there is a difficulty of knowing for certain that we have an “up and down” sounding, and the hardening of the “arming” by the cold and pressure, unless we bring up something we cannot be sure that we have touched the bottom; leaving the doubt on this point on one side, unless we use a very heavy sinker, so as to get an indication of the released strain when it touches the bottom, we encounter another complication.

Sir William Thomson’s sounding wire has added the element of reliability to our soundings in this latter case.  The note given out by the wire when the bottom is reached is perceptibly different when under strain, even if the dynamometer should give an unreliable indication.

It has been found that when a “bottom” has been recovered by the arming with tallow, the adherent grease seriously detracts from the value of the specimen for scientific purposes.  Washing with perfectly pure bisulphide carbon will save the sounding, but of course any living organism is destroyed.  As we have plenty of contrivances for bringing up loose “bottoms” without arming, we have nothing to fear on this score.