November 29, 1901.] 



SCIENCE 



857 



streams with lines parallel to one or another of 

 these fault systems has led Hobbs to infer a 

 very general dependence of stream courses on 

 fault lines or on faulted troughs (' The River 

 System of Connecticut,' Joum. GeoL, IX., 1901, 

 469-485). He notes that the modern school of 

 physiographers attach ' little importance to 

 geological structure planes as a factor in deter- 

 mining the position and the orientation of 

 water courses ' ; and if it is joint and fault 

 planes that are meant by ' structure planes ' 

 this comment is probably deserved as far as 

 explicit mention of such controls is concerned. 

 Nevertheless, the value of faulting in initiating 

 surface inequalities, and hence in determining 

 the course of consequent streams at the time of 

 faulting, as well as in producing planes of struc- 

 tural weakness, and hence in determining the 

 later development of subsequent streams, has 

 not been altogether overlooked. But it is 

 under the interaction of many controls that 

 modern physiography finds explanation for 

 river courses, and it is the want of considera- 

 tion of other controls than faults that leaves 

 Hobbs' paper unconvincing. 



The author extends the systems of fault lines 

 from the small Pomperaug basin all over Con- 

 necticut, preserving them rigidly straight and 

 parallel throughout. He then compares these 

 lines with the stream courses as shown on 

 the topographic map of the State, and where 

 a fair coincidence is found he concludes that 

 there is a relation of cause and effect, thus 

 explaining ' the definite orientation of water 

 courses.' There are serious difliculties in the 

 way of accepting this conclusion. 



It is inherently improbable that the Pom- 

 peraug fault lines possess an extension all over 

 the State in systems so rigid as are here postu- 

 lated. Some one of the infinite varieties of 

 curvature is much more probable than the 

 highly specialized case of straight paths. Strict 

 parallelism is also improbable as compared to the 

 many possible patterns of more or less distinct 

 divergence. The possibility of accidental coin- 

 cidence between eight systems of lines, seven 

 of which run between S. 48° W. and S. 44° E., 

 and the river courses of a region whose slope is 

 southerly and whose structural features (inde- 

 pendent of faults) frequently have a similar 



trend, is not given sufficient consideration. 

 Moreover, the branchwork pattern of the Con- 

 necticut valleys has little resemblance to the 

 network pattern of valleys that have been worn 

 along ancient fault lines in a well-faulted area : 

 the Stockholm district of Sweden, for example. 

 This typical district shows a most characteristic 

 oblique reticulation in the arrangement of its 

 valleys, such as would be expected in a maturely 

 dissected upland, obliquely criss-crossed by 

 many fault lines ; and it has numerous isolated 

 lozenge-shaped uplands occupying the meshes 

 in the netwoi-k of fault-line depressions. In 

 Connecticut the valleys have not a reticulated 

 pattern, and the uplands as a rule are not loz- 

 enge-shaped, but- give forth spurs between rami- 

 fying valleys. Well defined as is the fault-line 

 network in the Stockholm district, its lines are 

 neither straight nor in groups of parallels ; they 

 exhibit just such irregularities of curvature and 

 divergence as might be expected in a geolog- 

 ical instead of in a geometrical design. 



While faulted troughs (Graben) might have 

 guided many Connecticut streams for a time after 

 the Triassic deformation, it is highly improbable 

 that such courses could persist during the deep 

 denudation that the region as a whole has 

 since suffered, in the course of which many 

 streams would presumably desert their original 

 courses in the process of adjustment to new- 

 found structures. While the brecciated belts of 

 fault planes might during the denudation of 

 the region frequently gain the patronage of sub- 

 sequent streams, all other belts of weak struc- 

 tures would be active competitors for such 

 patronage ; yet no account is taken of such 

 competition, although it must have been com- 

 mon. The northward course of the Farming- 

 ton, for example, follows the weak beds of the 

 lower Trias, and is oblique to the known faults 

 of its district. Many of the existing streams in 

 the lowlands have courses consequent upon the 

 form of the glacial drift ; such is the case with 

 the Quiunipiac below Meriden ; and if some of 

 these streams have now cut down here and 

 there to bed rock, it is only by the chance of 

 superposition that the rock is found, as in the 

 Connecticut above Hartford. The possibility 

 that the lower courses of the Connecticut and 

 Housatonic have gained their southeast trends 



