ALTERNATIVE HYPOTHESES 397 



Other objections are that the syenites greatly surpass the ijolites in 

 volume, and that their concentric arrangement about the center is, to 

 the best of my knowledge, quite unparalleled in dikes elsewhere ; also, 

 according to Williams' view, we should expect to find the basic syenites 

 beneath the " hornsto'ne " zone, but, on the contrary, it is always syenite, 

 either leucitic or nephelinic, which crops out through this, as is seen on 

 Williams' map. 



The few instances mentioned by Williams* to prove that the syenites 

 are later than the ijolites and the monchiquitic dikes are of small mo- 

 ment and not conclusive, since they are just as easily accounted for as 

 last injections of still fluid syenite magma in the cracked complex. 



Apart from Williams' view of two distinct periods of intrusion (be- 

 sides that of the monchiquitic dikes), which we have seen to be unten- 

 able, the only alternative to differentiation of an originally homogeneous 

 mass which occurs to me is the " osmotic hypothesis" of Johnston-Lavis.f 

 This supposes the variation in composition of a rock mass to be due to 

 interaction of the magma with the conduit rock. The surrounding rocks 

 here, as far as we know them, are. either sandstones or the more abun- 

 dant shales. The composition of a similar shale from Little Rock J is 

 given here : 



Si0 2 = 56.30, A1 2 3 = 23.39, Fe 2 3 = 9.29, MgO = 1.49, CaO=0.36, 

 Na 2 = 2.76, K 2 = 1.36, H 2 == 5.16, FeS 2 = 0.26 ; sum = 100.37. 



It will be evident later that these shales could not furnish enough 

 silica and alkalies to " acidify " (feldspathize) the first and basic part of 

 the magma in a way corresponding with the explanation given by John- 

 ston-La vis in the case of Square Butte. § Indeed, any such explanation 

 based on successive upwellings of material seems to be quite out of the 

 question, and, inasmuch as such shales, or else sandstones or novaculites, 

 form all the country rock of this part of Arkansas, Johnston-Lavis' hy- 

 pothesis must be regarded as untenable here. 



Since erosion and stream action have not cut deeply into the complex, 

 we have no knowledge of its lower part. We do not know whether 

 there is a true floor, as in the typical laccolith, or whether the mass of 

 igneous rock extends downward for an indefinite distance as a stock. 

 We can only be certain from the contacts at the borders and from the 

 presence of the ridge of metamorphosed rock that the igneous complex 

 which is visible is the uppermost portion of a mass. For this reason 

 any section would be largely hypothetical, so none is attempted here. 



♦Williams : Op. eit., pp. 174, 188, 342. 



f Johnston-Lavis : Nat. Sci., vol. iv, 1894, p. 134. 



J Williams : Op. eit., p. 263. 



§ Johnston-Lavis : Report Brit. A. A. S., 1896. 



L VII— Bull. Geol. Soc. Am., Vol. 11, 1899 



