354 NATURAL SCIENCE. 



JULY 



still communicate with the rivers tend to become completely isolated. 

 In fact, the size of the Broads may be some measure of their age : 

 as the silting commenced at the seaward end, the largest Broads 

 would be expected to occur at the upper parts of the valleys. Such, 

 indeed, is the case ; thus the great Hoveton and Wroxham Broads on 

 the Bure, Barton Broad on the Ant, Hickling Broad on the Hundred 

 Stream, Surlingham Broad on the Yare, and Filby and its associated 

 Broads on the Muck Fleet, are all in the upper part of the valleys. 

 A gradual decrease can be traced as the Broads are followed towards 

 the sea, of course excluding those off the main courses of the rivers : 

 they become smaller, more regular in outline, completely isolated, and 

 further removed from the rivers. Thus on the great flat plain through 

 which the Ant flows south from Barton there are two small isolated 

 Broads, one a mile to the west of the river, and the other half a mile 

 to the east. Along the lower courses of the rivers there are 

 numerous such remaining to bear witness to the great sheets of waters 

 of which they are the diminished representatives, much as the small 

 oceanic atolls remain to attest the presence of the extensive fringing 

 reefs from which they have arisen. 



The peculiar type of scenery of the Broad area is thus the 

 outcome of exceptional physical conditions ; but we are tempted, 

 in conclusion, to inquire whether there is any geological evidence 

 as to the occurrence of similar conditions in the past. In considering 

 this point it may be advisable to attempt to form some idea of the 

 evidence that would be preserved in this area upon which a future 

 geologist would have to rely in trying to picture its present physio- 

 graphy. To do this let us return to Horsey, and note how the 

 north-easterly winds are driving the sand dunes over the adjoining 

 country. A section here would expose a basement bed of clay 

 containing estuarine and freshwater fossils, with occasional bands 

 of marine shells, and also the roots of aquatic plants or of trees that 

 love a damp soil. Above the clay would be a seam of vegetable 

 matter formed of the stems of similar plants, with occasional clumps 

 of tree stems : the whole of this seam would be remarkably free 

 from mechanically derived sediment, thanks to the sieve of bulrush 

 stems that grew around the margin ; this bed would be in turn 

 covered by a mass of sand or sandstone. But this is simply the 

 description of a coal seam with its root-bearing underclay and its 

 sandstone roof. Further, the occasional submergence of the coal- 

 field, as indicated by the layers of estuarine or marine shells, the 

 general flatness of the area (upon which Croll has laid such stress), 

 and the purity of the coal secured by a marginal filter of plant stems, 

 like the mangrove swamps of the Mississippi (to quote Lyell's com- 

 parison), are all points of striking resemblance between the Broads 

 and the conditions under which our coal-fields were deposited. 



Hence, though the Broads may not tempt ihe geologist with 

 any large series of instructive diagrammatic sections, nor yield to 



