THE NATURAL HISTORY OF IGNEOUS ROCKS. 207 



districts of regular sill-formed intrusions is in accord with 

 this. The sills themselves have presumably been fed by 

 dykes, but these are in the nature of the case rarely ex- 

 hibited. Again, the sill form seems to be related also to 

 the composition of the magma so injected. It has been 

 remarked by more than one writer that typical sills are 

 usually of basic rocks, and this observation seems to con- 

 nect itself with the greater fluidity of basic as compared 

 with acid magmas. The thickness of the "cover" may 

 also perhaps influence the form assumed by an igneous 

 magma injected among stratified rocks. Russell (15) gives 

 reasons for believing that sills have in general been formed 

 at no great depth below the surface at the epoch of the 

 intrusion ; and this he explains by the consideration that, 

 the less the energy expended in lifting the overlying rocks, 

 the more will be available for effecting lateral expansion. 

 There must of course be a limit to the operation of this 

 rule, since if the pressure above be insufficient to restrain 

 the fluid magma, the latter will break out. What this limit 

 may be, we cannot at present state ; but certain observations, 

 to which reference has been made in a former number of 

 this journal, 1 indicate that under submarine conditions it 

 may be very small. Certain basic sills in Cornwall and in 

 California, showing curious curvilinear divisions, have been 

 supposed to represent injections effected among deep-sea 

 sediments very near to the ocean-floor. Geikie (13) has 

 recently pointed out the wide distribution of this " pillow- 

 structure" amono- certain basaltic and andesitic rocks in 

 Britain, especially in the Ordovician of the Ballantrae dis- 

 trict and Forfarshire, and of Tyrone, Lough Mask, and 

 other parts of Ireland. He ascribed it to the lavas having 

 flowed into water, but in some instances {e.g., at Cader 

 Idris) the hypothesis of intrusion seems also to be a tenable 

 one. In most cases associated radiolarian cherts give evi- 

 dence of thoroughly deep-sea conditions. 



These various points come out more clearly when the 

 characteristics of sills are compared with those of laccolites. 



1 Science Progress, vol. v., p. 480. 



