Some European geomorphologists now treat "ria" as merely a synonym of "drowned valley," without distinction as to kind or s ze, including some mere creeks in this category. Thus, in a recently published work on coastal geomorphology by Guilcher, "ria" is defined as any valley system partly invaded by the sea. "Most commonly the drowning is due to the Flandrian transgression; it can also be the result of a subsidence" (Guilcher, 1954, pp. 117, 119). This practice follows sugges? tions of Gulliver (1899, pp. 167, 220) and the usage of Johnson (1919, p. 173), who described all coasts of submergence with the exception of fiord coasts as "ria shorelines," and it may be defensible to some extent on the ground that if such a definition of "ria" is adopted the use of the word facilitates concise statement and euphonious expression. As de Martonne (1935, p. 1023) has pointed out, however, the usage cannot be regarded as correct. Nor does it seem desirable, apart from academic considerations of priority, to extend so far the meaning of a term already usefully em? ployed more or less exactly in the sense in which it was understood and defined by Richthofen. Freely translated, Richthofen*s definition is as follows: "On coasts that are trans? verse [to the structural grain of the land] the ends of ridges commonly jut out boldly into the sea, while [in between the promontories so formed] the sea invades the lower reaches of the re-entrants and valleys that separate ridges. Highly irregular coastal outlines are the rule, generally fringed by outlying islands. Such rias coasts are pro? duced by positive shift of the strandline. . . . Though these [structural] conditions are only imperfectly fulfilled on the Galician-Asturian coast, it is only there that arms of the sea characteristically associated with them have a specific name (rias); the same designation may be applied to the more perfectly developed forms of other regions." (Richthofen, 1886, pp. 308-9.) Thus, adopting ria from the names given to gulfs or wide-mouthed bays with considerable extension inland in north-west Spain, Richt? hofen established the geomorphic use of "ria" as descriptive of some bays and gulfs similar in a general way to these. He assumed that the word, used as an empirical geographie term, already had that meaning; but he defined it more strictly in relation to "rias coasts;" and this redefinition holds good in geomorphology notwithstanding the application of the same name in Spanish to any river mouth. (Compare the artificially restricted geomorphic significance of "cuesta," another Spanish word.) Penck (1894, ii, p. 566) has supplied a generalized description of typical Galician gulfs mapped as rias (Ria de Vigo, etc): they reach to a maximum distance of 30 miles inland; they are funnel-shaped, or widen seaward, i.e. few or no parts are landlocked; and they are without relation to structure in the terrain (which is granite). The last-mentioned characteristic, indifference to structure, was not included by Richthofen in the definition of "ria" as a geomorphic term applicable to some other regions. A coast characterized by rias, or "rias coast," is on the contrary, as defined strictly by Richthofen, the deeply drowned margin of a terrain (generally of folded stratified rocks dissected by subsequent valleys) with trend or strike transverse to the