370 TMRNARIl AXn l,AnT) [fHAr. 15 



are particularly iiicoini)lete in the Atlantic and Indian Oceans where they are 

 based on ])uhlished soundings rather than on original and un])u})lishod sound- 

 ings as they are in the Pacific. Therefore, the charts should not be inter])reted 

 to indicate that seamounts definitely are more common in the Pacific than 

 elsewhere. New seamounts are invariably discovered \\henever an oceano- 

 graphic ship crosses a previously unsounded region in the Pacific. This circum- 

 stance may be used as the basis for a prediction of the total number of seamounts 

 that would be found if the whole Pacific were sounded along closely spaced 

 lines (Menard, 1959). It is assumed that the seamounts of all sizes are cones 

 Avith uniform slopes and that, in general, they are randomly distributed. No 

 significant error is introduced by the former assumption, but the latter limits 

 the size of the region that is studied because some topographic provinces have 

 more seamounts than others. The Baja California Seamount province between 

 Mexico and Hawaii was selected for study and all the seamounts w^ere counted 

 along 7400 km of sounding lines. About 10^ seamounts with at least 1 km of 

 relief appear to lie within the area of 1,400,000 km^ or about one per 1400 km^. 

 Extrapolating to the whole Pacific Basin and considering various sea-floor 

 topogra])hic provinces, the total number would be roughly 10^. A second 

 method of estimating is to count all the seamounts in areas that are surveyed 

 densely enough to find every seamount and to extrapolate to the remainder of 

 the basin. The same number of 10^ results. Extrapolating to the entire world 

 ocean is more uncertain, but if the concentration per unit area is constant in 

 any large region, the world total is about 2 x lO-*. 



A typical large example is Gilbert Seamount in the Gulf of Alaska (Fig, 3). 

 It consists of a roughly circular main cone with a diameter of 20 mi, an eleva- 

 tion of about 11,000 ft, and a smaller parasitic cone on one flank (Menard, 

 1955). The side slopes are slightly concave upward and reach a maximum of 

 22°. The maximum slopes are about the same on smaUer seamounts but. in 

 general, the slopes do not appear to be concave upward if the peaks are less 

 than one mile high. Some seamounts are more elongate with a length as much 

 as four times the width ; in this characteristic, they resemble the shield volcanoes 

 of the Hawaiian Islands, wdiich tend to be elongate because they lie on linear 



rifts. 



Most seamounts are clustered in lines or in elongate groups of roughly 10-100 

 and consequently, as groups, they closely resemble the volcanic archipelagoes 

 of the Pacific Basin such as the Samoan and Hawaiian Islands. This is charac- 

 teristic of the seamounts of the Pacific, but at least one prominent linear group, 

 tbc Kelvin Seamounts (Heezen et al., 1959), lies in the Atlantic. Most of the 

 linear groups of seatnounts of the northeastern Pacific have the same north- 

 westerly trend as the linear island groups of the central and western Pacific. 

 Likewise, the individual seamounts in a cluster are connected in various ways 

 like archi])elagoes. The Mid-Pacific Mountains (Hamilton, 1950) rise from a 

 ridge with steep sides like the Tuamotu Archi])elago. Other seamount clusters 

 such as the outer ones in the Gulf of Alaska (Menard and Dietz, 1951) stand 

 above low rises — resembling the Marquesas Islands in this respect. Some 



