FORCES AND PROCESSES IX OCEAN BASINS 49 



the movement of India away from Africa and with an incipient 

 sphtting of Africa from Arabia (Girdler, 1958). Such an explana- 

 tion is not, however, available for the ridges of the Pacific which 

 are supposed to be connected with the ridges of the Indian Ocean 

 and thus with the mid-Atlantic ridge. It is important to be sure 

 of the various suggested connections and to know how far the 

 ridges are really similar. In particular, do the Pacific ridges often 

 or usually lack a central valley as has been suggested by Alenard 

 (1958)? 



It has been suggested (Hope, 1959, summarizing Russian \\ork) 

 that the Lomonosov ridge which crosses the Arctic Ocean is a 

 folded feature connecting the \'erkhoyansk range of Siberia to 

 Ellesmere Island. Obviously the ends of this ridge could only 

 remain attached to Asia and to Ellesmere Island while North 

 America moved through 70° in longitude if it were considerably 

 distorted. It therefore seems of great interest to examine the 

 materials and date of this ridge and the form of the connections 

 with the continents at both ends. 



If the continents are moving, there must be horizontal forces, 

 and the only plausible suggestion that has been made is that these 

 forces are associated with convection currents in the mantle. The 

 high heat flows found on the mid-Atlantic ridge and on the east 

 Pacific rise suggest that, if the currents exist, their rising limbs are 

 under the midocean ridges. The rising current would be of lower 

 density than the material on each side and would explain the 

 rough isostatic compensation of the ridge. Hess (1954) has also 

 provided an alternative explanation depending on serpentization 

 of material beneath the ridge which does not appear inconsistent 

 with a rising current beneath it. 



Obviously, views about the existence of convection currents 

 and continental drift must be tentative. The mechanisms, if they 

 exist, are concealed so effectively and have such long time scales 

 that they cannot be directly observed. Theory cannot assert or 

 deny the possibility of such processes. If the facts show a high 

 probability that they have occurred, we must make the best 

 theory we can, but we should not place too much reliance on it 

 or let it obscure the very complicated facts. 



