Vol. 50.] CERTAIN NOVACULITES AND QUARTZITES. 383 



Reverting once more to the rhombohedral cavities which occur in 

 Ouachita stone, there seems to he an additional reason for believing 

 that they were originally occupied by dolomite and not by calcite, 

 from the fact that the rhombohedron, when uncombioed with other 

 forms, is not common in calcite, while it is of extremely common 

 occurrence in dolomite. 



The general absence of fossils in the actual novaculites of 

 Arkansas, although some, chiefly the ossicles of crinoid-stems, have 

 been met with, 1 need form no barrier to the hypothesis that the 

 rock, is a siliceous replacement of a limestone, since it is common to 

 find beds of dolomite almost or quite destitute of fossils, and, as 

 Prof. Renard remarks, " gelatinous silica moulds itself upon objects 

 and preserves their forms ; dolomite, on the contrary, through its 

 tendency to develop terminated crystals, tends to efface them when 

 it becomes infiltered among organic remains." 2 Furthermore, the 

 stratigraphical relations of the Arkansas novaculites are not incom- 

 patible with the assumption that they are the siliceous replacements 

 of limestone ; for, according to Mr. Griswold, they occur between 

 shales of late Lower Silurian 3 (Ordovician) age in which graptolites 

 are found, including the genera Diplograptus and Dicranograptus, 

 both of which are abundant in the Bala series of Wales, while the 

 former genus, if not the latter, occurs also in the black graptolitic 

 shales of the Coniston Limestone series. 



The association of the Arkansas novaculites with black grapto- 

 litic shales, and that of our own Ordovician limestones with similarly 

 coloured graptolitic shales, is not, certainly, a proof that the nova- 

 culites are representatives of Ordovician limestones, but it is a 

 coincidence which, to say the least, is significant. 



At this stage it seems desirable to discuss the character of the 

 siliceous grains which constitute the Arkansas novaculites. 



In a paper by Messrs. A. J. Jukes-Browne and W. Hill 4 remarks 

 by Dr. G. J. Hinde are quoted, in which he says that he " is unable 

 to explain the causes which have produced this singular [globular] 

 form of colloid silica, or to say why the silica of the sponge- 

 spicules should not have passed into the more stable condition of 

 chalcedony or crystalline quartz, as is the case with those of most 

 other fossil sponge-beds." Here we have Dr. Hinde's statement 

 that sponge-spicules may be found not only in the condition of 

 colloid silica (opal), but also in the cryptocrystalline condition 

 (chalcedony) and in the crystallized condition (quartz). This appears 

 to indicate that the originally colloid silica of a sponge-spicule may 

 undergo a series of changes which culminate in its conversion into 

 quartz. May not that which happens in a sponge-spicule also 

 occur in the silica which replaces a limestone ? 



1 ' Whetstones and the Novaculites of Arkansas,' p. 133. 



2 ' Des Caracteres distinctifs de la Dolomite, etc.,' Bull. Acad. roy. Bel»ique, 

 ser. 2, vol. xlvii. (1879) p. 562. 



3 ' Whetstones and the Novaculites of Arkansas,' p. 205. 



4 'The Occurrence of Colloid Silica in the Lower Chalk of Berkshire and 

 Wiltshire,' Quart. Journ. Geol. Soc. vol. xlv. (1889) p. 407- 



2d2 



