354 



SCIENCE 



[N. S. Vol. XXVI. No. 663 



rent. Otago peninsiila would thus in several 

 respects resemble Banks peninsula, on the 

 same coast farther north; for this is again a 

 -dissected and formerly insular volcanic mass 

 with a ragged and cliffed outer shore line, 

 now transformed into a peninsula not only 

 by the flying northeast stretch of Ninety-mile 

 beach, but apparently also by the forward 

 growth of the fluviatile Canterbury plains in 

 the sheltered waters back of the former island. 

 Marshall's account of Otago peninsula pro- 

 ceeds on the " two bites of a cherry " method 

 of first describing the various surface features, 

 and then explaining their origin. So cautious 

 a method may be appropriated in treating land 

 forms of uncertain origin ; but its employment 

 in so simple a case as this one would seem to 

 indicate an undue consideration for those who 

 even in this day need to have it explained that 

 bays are half-drowned valleys. Much space 

 might be saved if the peninsula were briefly 

 described as having been submaturely dis- 

 sected in a former cycle of normal erosion, 

 and as, after a depression of towards a thou- 

 sand feet, being now vigorously attacked by 

 the sea on the new shore line which is already 

 advancing towards maturity on its seaward 

 side. It is a great advantage to the reader 

 to have the essence of the story thus presented 

 in condensed form at the outset; the details 

 can then be easily apprehended in their proper 

 relations as they are reached iji further read- 

 ing. 



THE FAYUM DEPRESSION, EGYPT 



Imagine a series of strata, of which certain 

 members, J to P, are 700 meters in thickness, 

 dipping very gently to the northwest. Let 

 the lower formation, J, be a resistant lime- 

 stone, 30 m. thick; the next formation, K, a 

 series '.of weak clays and marls, 70 m. thick; 

 and the following members, L to P, a succes- 

 sion of alternately resistant and weak strata, 

 480 m. thick. Conceive the whole series worn 

 down nearly to baselevel in a desert climate, 

 thus producing a broad peneplain on which 

 the beveled strata appear in belts trending 

 northeast-southwest. Now let the peneplain 

 be uplifted with a gradual slope to the north, 

 so as to gain an altitude of 300 or 400 m. in 



the district here especially considered; and in 

 consequence of this uplift imagine the barren 

 surface to be dissected to a stage of maximum 

 relief by the winds and occasional raiiis. The 

 weak belt, K, will thus be irregularly exca- 

 vated as a subsequent depression along the 

 strike of the guiding formation; the depres- 

 sion will be bordered on the southeast by a 

 structural or " dip " plain of the underlying 

 limestone, J; and enclosed on the north by 

 three cuestas, rising in ragged escarpments, 

 L, N, P, and separated by broad steps, M, 0. 

 The upland beyond the highest escarpment will 

 gradually descend far northward to the sea, 

 younger and younger formations being crossed 

 on the way; while in the opposite direction the 

 rising plain of the underlying limestone 

 cuesta, J J will presumably break off in a south- 

 east-facing escarpment overlooking another 

 subsequent lowland eroded on underlying 

 strata; . . . and so on to the basement old- 

 land. 



The waste from the depression eroded on 

 the weak belt, Ej having been largely exported 

 as dust by the winds, the floor of the de- 

 pression will sink here and there in enclosed 

 basins, which may be excavated even below 

 sea level; and the basins will be separated by 

 low residual portions of the weak strata, which 

 will form what may be described as transverse 

 barriers or ridges — there being as yet no tech- 

 nical name for such features. Along the east- 

 ern side of the district, toward which the up- 

 lifted peneplain may have had a faint slope, 

 imagine an additional uplift by faulting or 

 monoclinal bending; and along the trough 

 thus defined let a large north-flowing river 

 erode a mature valley through the desert. The 

 western side of the valley will vary in height 

 as it obliquely cuts the several cuestas; with 

 the eastern side we are not especially con- 

 cerned. While the main valley is worn down 

 contemporaneously with the general dissection 

 of the peneplain, the river happens, by lateral 

 erosion, to wear through the first transverse 

 barrier in the weak beds, K, that separates the 

 valley from a neighboring subsequent basin; 

 a branch of the river then flows into the basin 

 and forms a lake ; but as the river continues to 



