part 4] .TUBA881C CHEONOLOGY. 415 



disorder ; ((/) that the second white-stone deposit has practically 

 come into contact with the first one, giving the Watton Bed a 

 false appearance of stratal continuity, whereas it masks a con- 

 siderable non-sequence — say, some dozen or more hemeriB ; (e) that 

 the second white-stone deposit which is of two dates at Watton 

 Cliff — Hammatoceras and earliest Dumortieria — has been partly 

 removed, so far as the Dumortieria portion is concerned, from 

 Bothenhampton, and has been wholly removed from the Thorn- 

 combe area : there, indeed, removal has gone further — it has taken 

 off the deposit of striatulum date from Thorncombe Beacon, and 

 yet left it as witness for original deposition at the neighbouring 

 Doghus and Down Cliffs, in places. 



The phenomenon of what may be called stratal repetition — 

 the occuri-ence of like deposits after greater or less intervals of inter- 

 ruption by deposition of strata unlike them — is quite well known. 

 It occurs in two 'forms which may be called local and non- 

 local — the first is stratal repetition in the same district, the 

 second is stratal repetition, but not in the same district ; there is 

 stratal repetition in time, although what may be called the 

 de positional focus has shifted. In the first case the like 

 deposits are superinqDosed, in the second case the}^ are not. 



Those Ju)-assic sands of the South- West of England which 

 makes a first appearance in Middle Lias (Domerian), and continue 

 with various changes of locality until early Inferior Oolite 

 (Aalenian), illustrate remarkably well stratal repetition, both local 

 and non-local, as Table III (p. 415) will show. 



Before illustrating this point I ma}^ sa}^ just a word about the 

 so-called 'Midford Sands' or 'Oolite-Lias Sands' (Cottes wold, Mid- 

 ford, Yeovil, etc. Sands). From what one may read concerning 

 lithological [lithic] evidence about change of deposit producing 

 change of ammonite fauna, and when one sees assumptions made 

 that the same formation stretching across country may be taken 

 as evidence for the same date, I fear that the lessons to be 

 gathered from the ' Midford Sands,' the secret of which I un- 

 ravelled some years ago,^ have not yet been learnt. Here is a 

 formation of sand stretching across country for some 90 miles, 

 from near Gloucester to the Dorset coast, resting upon clay below 

 and capped by limestone rocks above. But the sand formation 

 is not uniform in date — as it passes from north to south it 

 gradually becomes later in time. Diagram 10 (p. 414) will illus- 

 trate this — the numbers refer to the hemerse in Table III ; but, 

 of course, the diagram does not illustrate all the complexities due 

 to differential thickening of deposits, to penecontemporaneous 

 erosions, and so on. But it shows how unreliable a guide the 

 lithic character may be. In this case clay, sand, and limestones are 

 all being deposited synchronously during several hemerse without, 

 so far as is known, any effect on the ammonite faunas. But it can 



1 ' On the Cotteswold, Midford, & Yeovil Sands, &c.' Q. J. G. S. vol. xlv 

 (1889) p. 440. 



2f2 



