TO W. GARNETT, Esq.

                                                                     Glenlair, 11th August 1877.

     Your experiments on electrified paraffin oil are excellent, and may lead to increase of knowledge.

     If the fluid dielectric and also the air are perfect insulators, nothing can get electrified, but the equation at the
     surface, instead of being P= Po, will be
 
 

     (excluding capillary action) where   is the resultant electric force normal to the surface and just outside it.
     This causes the surface to rise wherever the normal force is great, or close to the electrodes.

                                             · · · · · ·

     The science of displacements is in Euc. I. 4, etc., and wherever one figure is placed upon another. It belongs to    the method of contemplating the relations of two figures which may be supposed to co-exist, though we may also   suppose that they are copies of the same figure in different positions.

But just as we assume that distance is a continuous quantity capable of measurement, though all our attendants at    measurement are made with instruments made of non-rigid and discontinuous matter, so we may assume that time   is a continuously flowing quantity capable of measurement, though we have not yet found out any accurate method  of comparing distant intervals of time.

     Now Kinematics requires no more than this notion of time, as the common independent variable t. If we suppose     that is that (unknown) which flows uniformly then for kinematical purposes it is enough that t is a function of;   but    when we come to Kinetics proper we must have    very small.

     Have you read Julius in Nature, about the beginning of June? [14th June].

     The most constant things we know are the properties of bodies. For instance, water in equilibrium with ice and   vapour gives us a good deal.

     I. A unit of density (not the orthodox one) .

     II. A unit of pressure (too small for practical use) .

     III. A unit of time (namely, the time of revolution of a satellite just grazing a sphere of water) = T.

These three quantities being independent of each other give M, L, and T. [401]

P/D gives a (velocity) squared which could also be got from the P/D of the vapour (a different one).

Then this gives also a standard temperature; all that we want is to get pure water.