1 80 Prof. Baker, Transformation of equations of electrodynamics 



these being the equations (228), p. 146 of M. Th. de Bonder's paper, 

 in a form which to the present writer at least seems much more 

 intelhgible. 



We shall then naturally take p, V/, V,/ , V/ so that 



p' {E^'V/ + ... + E,') = p {E^V, + ... + E,), 



or p' {E') {V) = p (E) {V), ^p'.p{E){r)^p{E){V) 



or p' (E) {pV) = wp {E) {V), 



namely, so that wp {V) = p (^F'), 



or p (jOiiF^''+ Vv^Yy'-V ^13 7^'+ p^^ = wpV^, 



P {VilVx'+ V^%yy'+ Pi3Vz'+ Pu) = ^P' 



which give, for instance, 



^x = {PiiV/+ ... + Pu)/{p^iV^'+ ... + p^), 

 and this, comparing with 



8x/8t = {p-^-i8x'+ ... + Pii8t')/{p^^Sx'+ ... + p^^St'), 

 is in accordance with the view which regards V^c as a velocity. 



Remark. We may evidently use Grassman's units in § 2 instead 

 of the quaternion units, the Maxwell equations being a particular 

 case of those considered here. 



