1918] Davis: The Radiolarian Cherts of the Franciscan Group 335 



from rocks similar to the cherts and shales of the Monterey group of 

 California. 00 



Griswold stated that the novaculites contained no organic remains, 

 save at one locality where crinoid stems were found. Recently Charles 

 Lawrence Baker, of the Bureau of Economic Geology of the State of 

 Texas, has discovered radiolaria in these rocks. Through his courtesy 

 the following description of the novaculites of Texas is presented : 91 



The Caballos novaeulite of the Marathon basin of Trans-Pecos Texas appar- 

 ently occupies the same stratigraphic position as the Arkansas novaeulite of the 

 Ouachita Mountain region of west -central Arkansas and east-central Oklahoma. 

 In their respective regions both novaculites unconformably overlie strata of Fern- 

 vale-Richmond age ; both are strongly folded with the associated sedimentary 

 rocks in the structures of the Hercynian diastrophism, apparently contemporane- 

 ous in the two regions. E. O. Ulrich correlates the two novaculites and places 

 them in the Oriskany, or upper part of the lower Devonian. 



The lower forty feet of the Caballos novaeulite is rather thin-bedded light 

 brown chert with some layers of white novaeulite. The upper fifty feet is massive- 

 bedded, ripple-marked, and much-fractured white novaeulite. No interbedded 

 shale occurs as in the case of the Arkansas novaeulite. Under the microscope the 

 commercial varieties of both Arkansas and Caballos novaeulite, suitable for use 

 as fine abrasives, consist simply of a mosaic of interlocking, very fine and uniform 

 grained, angular-contoured quartz particles. The non-commercial varieties, on the 

 other hand, much more resemble the ordinary run of cherts. Under the micro- 

 scope the latter variety exhibits irregular, somewhat rounded aggregations of 

 coarser-grained quartz particles with wavy extinction in a ground mass of finer 

 and more uniform grained irregular-contoured quartz particles arranged in a 

 mosaic. The coarser aggregations or knots are sometimes seen to be the imper- 

 fectly preserved remains of radiolarians, each knot being the remains of one radio- 

 larian. Only one specimen of the non-commercial variety of the Arkansas novaeu- 

 lite was collected by the writer. A slide made from this showed no radiolarians 

 but precisely similar structure as the non-commercial varieties of the Caballos 

 novaculites. 



Both the Arkansas and the Caballos novaculites exhibit broad ripple-marking. 

 There may be some question whether these are really ripple marks. Conceivably 

 they might be produced by the Hercynian folding. The commercial varieties, 

 which do not exhibit traces of radiolaria, may have been an original siliceous ooze 

 in which the amorphous, soluble silica of the radiolarian skeletons had been 

 changed to quartz before final consolidation, or selective metamorphism may have 

 destroyed the radiolarian structures after consolidation, or there may never have 

 been any radiolarians present in the commercial varieties. It is perhaps signifi- 

 cant, as regards origin, that the radiolarian remains in the non-commercial 

 varieties of the Caballos novaculites are always partially destroyed. 



The Caballos novaeulite is overlain, probably unconformably, by the Santiago 

 chert, in some respects an even more remarkable formation than the novaeulite. 

 The Santiago chert is thin-bedded, banded, or ribboned, of dull shades of prac- 

 tically every color, but mostly green. Throughout its maximum-observed entire 

 thickness of 450 feet there is not even the thinnest layer of anything but chert. 



»o Jour. Geol., vol. 6, p. 371, 1895. 



si Personal communication, March, 1917. 



