368 
DES. T. ANDERSON AND J. S. FLEET ON THE ERUPTIONS OF THE 
beneath, but this is local. On the windward side, however, it is the exception to see 
fresh rock exposed on the surface of the terraces. So advanced is the decomposition 
that it is often very difficult to distiiiguish a weathered lava from an ash bed. A. 
brown earth covers everything, fine as a whole, but often full of rounded stones, 
which are usually the bombs and ejected blocks of the tuffs, more resistant to deca}' 
than the finer matrix between. Occasionally a lava assumes this ajDpearance, and 
then it is only by proving that all the rounded masses have the same petrogTaphical 
character, or Ijy finding that the matrix shows the same porphyritic crystals as the 
harder kernels it incloses, that we can establish what the rock was originally. The 
presence also of the red layer of baked soil beneatli the lavas often confirms the 
diagnosis. The more basic lavas weather spheroidally, and the brown, rotten earth 
shows then innumeralde spheroids, large and small, composed of concentric shells. 
The weathered material is known in St. Vincent as “ Pozzuolana,” and is much used 
for dressing the surface of the roads. 
On the coast, a little south of Georgetown, a pecidiar brown earth covers the 
volcanic rocks to a considerable depth. The road just north of Colonarie Point passes 
beneath cliffs of this material, over 50 feet in height, which continue up to near 
Black Point. It is dark-yellow or brown in colour, soft, fine-grained, with numerous 
l^locks and rounded fragments scattered through it. Occasionally it shows a fairly 
good bedding, the division planes being 2 or 3 feet apart. Some of the larger stones 
are vesicular volcanic Ijombs, others are pieces of various lavas, all of types frequent 
in the island. This dejDosit contains no fossils, and apparently has not accumulated 
beneath water, for there has been no sorting of the materials, and the rough imperfect 
bedding is quite unlike that of an aqueous formation. 
Its most significant feature is the dip of the rude bedding planes. When the 
material lies on slanting ground the dip is down the slope and parallel to the 
surface. On the rounded backs of the ridges between the valleys the beds lie flat. 
In other words, the bedding shows an intimate relationship to the surface configuration 
of the country. Not infrequently “ unconformabilities ” are to be observed (see 
Plate 22, fig. 2), slanting layers overlapping others nearly horizontal. The nature of 
the material above such an “ unconformability ” and below it does not differ. 
The origin of this formation is somewhat of a problem. In many features it 
resembles a pulverulent ash wdiich has been showered down from above, and has 
covered every irregularity of the surface with a layer of fine ejecta. Each bed might 
be supposed to represent an eruption, and the larger fragments might be bombs and 
ejected blocks. Such loosely aggregated ash might be expected to be especially 
liable to rapid decomposition. But a coating of this nature would certainly be readily 
eroded by the tropical rains, and washed off the slopes into the valleys. It should 
occur mostly near the Soufriere, as that is the only volcano in the island which shows 
any evidence of having been in activity for a long time past. This deposit, however, 
is not well seen on the Soufriere, but may often be found in patches near the south 
