TO H. R. DROOP, Esq. (of the Equity Bar). 

                                         Glenlair, Dalbeattie, N.B., 28th December 186 l. 

 I enclose a short statement of the scheme of endowing the chapel which was built near us in 1838 for this district, which is very far from any parish church. If we can raise £1000, there is a fund already raised which will   contribute £2000, so as to give a salary of £120 to the minister permanently, and as the people are too poor to   support the minister themselves, we hope to make the chapel independent of chance contributions in this way.   Great part of the funds for building the church were subscribed in London by all kinds of people who were friends of an English gentleman who then had property here; but we have no longer any such means of drawing on the   metropolis. 

If you can put us in the way of diminishing the deficit we shall be grateful, and I will see that the money goes to the  fund, and that the names are duly entered, however small the contributions. 

 . . . I have nothing to do in King's College till Jany. 20, so we came here to rusticate. We have clear hard frost    without snow, and all the people are having curling matches on the ice, so that all day you hear the curlingstones   on the lochs in every direction for miles, for the large expanse of ice vibrating in a regular manner makes a noise   which, though not particularly loud on the spot, is very little diminished by distance. I am trying to form an exact  mathematical expression for all that is known about electro-magnetism without the aid of hypothesis, and also   what variations of Ampère's formula are possible, without contradicting his expressions. All that we know is about    the action of closed currents—that is, currents through closed curves. Now, if you make a hypothesis (1) about   the mutual action of the elements of two currents, and find it agree with experiment on closed circuits, it is not    proved, for—- 

     If you make another hypothesis (2) which would give no action between an element and a closed circuit, you may    make a combination of (1) and (2) which will give the same result as (1). So I am investigating the most general    hypothesis about the mutual action of elements, which fulfils the condition that the action between an element and a  closed circuit is null. This is the case if the action between two elements can be reduced to forces between the   extremities of those elements depending only on the distance and + or - according as they act between similar or   opposite ends of the elements. If the force is an attraction 

                                      = (r) ss' (cos  + 2 cos  cos ') 

     where  is the angle between s and s', r the distance of s and s' and   and ' the angles s and s', the elements,   make with r, then the condition of no action will be fulfilled.