OCTOBBE 24, 1902.] 



SCIENCE. 



671 



boundary very correctly. In fact it is more 

 nearly correct than some of the later ones. 

 The beds which he refers to this period had 

 been previously noted by Shumard, Buckley 

 and others, and their age determined to some 

 extent by vertebrate remains found in their 

 upper portion. In 1894 I described these 

 beds * as they occur in southwest Texas and, 

 on the basis of Professor Cope's determina- 

 tions, separated the ISTeocene into Oakville 

 (Miocene), Lapara, Legarto and Eeynosa 

 (Pliocene). Later I traced these beds to east 

 Texas and proved their identity with Lough- 

 ridge's beds,t and thus found that the clays 

 and sands east of the Trinity, which Kennedy 

 has called the Fayette and Frio, are in fact 

 Oakville and Lapara-Lagarto. The only ex- 

 ception to this which I now recall is the 

 sandstone north of Corrigan, which Professor 

 Harris first thought was Lower Claiborne, but 

 after study of fuller collections decided to be 

 Jackson. 



Therefore the true correlation of the two 

 sections would probably be more like this: 



My interpretation would be that the Sabine 

 section shows an overlap of the Lower Clai- 

 borne on the Lignitic, entirely covering the 

 sandy, unfossiliferous Carrizo beds, which else- 

 where in Texas form so prominent a feature 

 at the top of the Lignitic beds. Also an over- 

 lap of the Jackson on the Yegua ? (Cocks- 

 field Ferry beds), covering both the Fayette 

 and the Frio. 



* Journal of Geology, Vol. II., pp. 549, etc. 



t Trans. Tex. Ac. 8c. , 1894, pp. 23. Trans. Am. 

 Inst. Min. Eng., Vol. XXXI. 



The Oakville is stratigraphically the cor- 

 relative of the Grand Gulf, and it is possible 

 that closer work in Texas may yet show that 

 the lower portion, in which we have found no 

 fossils as yet, is the extension of the Oligooene 

 portion of the formation. From Harris' de- 

 termination of the age of the Burkeville beds, 

 I suspect them to be a part of the Oakville 

 beds, as they are certainly older than any La- 

 para we know west of the Trinity. It will 

 require still further field work, however, to 

 determine its exact relation to these beds. 



E. T. DUMBLE. 



A NOTE ON METHODS OF ISOLATING COLON BACILLI. 



It often happens that bacteriologists wish 

 to obtain fresh cultures of Bacillus coli for 

 experimental purjaoses and they sometimes find 

 that the methods of isolation in general use 

 are unsuccessful or inconvenient. The rea- 

 sons for the latter fact have not hitherto, so 

 far as I am aware, been satisfactorily ex- 

 plained. In some comparative bacteriological 

 studies made in cooperation with one of my 

 students, Mr. William J. Mixter, I found it 

 necessary to obtain a large number of fresh 

 cultures of B. coli and soon learned that the 

 two methods in common use, viz. (1) 'plating 

 out ' the aqueous suspension of fresh faeces 

 in agar, litmus-lactose-agar, or gelatin, or (2) 

 inoculating from such a suspension into dex- 

 trose broth and incubating eighteen to twenty- 

 four lioiirs with subsequent plate cultivation, 

 while giving a plentiful supply of bacteria 

 gave, for the most part, negative results as 

 regards B. coli. 



After considerable experimenting we finally 

 hit upon the following method with satisfac- 

 tory results. A very small portion of fresh 

 fsEces is inoculated directly into dextrose broth 

 in the fermentation tube, and allowed to de- 

 velop at 37°. At the end of from two to six 

 hours the culture medium becomes turbid 

 throughout and gas formation is generally 

 proceeding rapidly. If inoculation is now 

 made into litmus-lactose-agar plates and incu- 

 bation continued at blood heat, colonies of B. 

 coli develop abundantly and with great rapid- 

 ity. Isolation, purification and cultural tests 

 can then be carried on by the usual methods. 



