My own field of seismology can tell us a little bit about how this process occurs. Using techniques analogous to those used in medicine, in which X-rays are used to image parts of the human body, seismologists can use seismic waves to image the interior of the earth. Here's an image on your left (fig.29) of the East Pacific Rise; an image on the right (fig.30) is part of the Mid-Atlantic Ridge. Warm colours are meant to depict slow seismic wave speeds, thought to be the result of high temperatures or perhaps even the presence of molten rock.
So these variations in wave speeds can be mapped out and more or less understood, but we're limited in the resolution with which we can do this kind of imaging. These two images have a resolution of about one kilometer. We can't see anything much finer in detail. And seismology is also limited in its ability to characterize in detail the composition of the rock, and to distinguish several important classes of rocks that are necessary to understand the process by which crust is formed and created.
I have several slides that I borrowed from Henry Dick , a scientist at the Woods Hole Oceanographic Institution , who makes the following case. The classical view of what the ocean crust is made of came from a combination of three different lines of evidence：
* There were, of course, rocks that were recovered from the sea floor by a ship dragging a device known as a dredge across scarps in the sea floor.
* There was seismic work that showed a layered structure for a particular velocity - the P wave velocity - versus depth.
* And there were exposed on land sequences of rocks that seemed to have the night properties to be analogous to the oceanic crust. These suites of rocks were called ophiolites and were thought to have been generated in the ocean at one time and then shoved through some faulting process onto land, where they are now exposed. And different ophiolite rock units have different thicknesses (fig.31).
But what Henry Dick and others have pointed out is that at least on many of the mid-ocean ridges, the proportion of rocks that are found on the sea floor is wrong for this model. In particular (if you're not a petrologist, don't worry about the names) this rock shown in fig.32 as green called peridotite, which is thought to be at the very bottom of the sequence and therefore should be relatively rare on the sea floor, is in fact the most abundant returned rock in dredge hauls from slow spreading ridges in the Atlantic and the Indian Ocean.
So that result has led to different ideas for the structure of fast spreading oceanic crust like that in the Pacific (fig.33), and slow spreading oceanic crust like that in the Atlantic (fig.34). And you see these differences in these cartoons from Henry Dick. There's lots of technical details, please don't worry about them too much.
But some key characteristics are that there's some pretty complex layering predicted in the lower crust and that the transition from the crest to the mantle may be quite gradual. Here you see an inter-fingering of crustal and mantle rocks extending over several kilometers. At the slow spreading ridges it is thought there is tremendous variation as you go along a ridge in the thickness and the characteristics of the crust . Despite the artist's skill, we must remember that we have not sampled directly anything but the very top of the oceanic crust. We've gone down a couple of kilometers in one place, but we have never sampled the middle or the lower crust and we've certainly never sampled the target Dr. Hirano mentioned - the Earth's mantle.
