OCTOBEB 4, 1907] 



SCIENCE 



451 



stream length that is reducible to definite 

 directions and with a statement of the amount 

 in tabulating the measures, is here much 

 needed. A quantitative discussion of coinci- 

 dence is also desirable. 



On the second point, there is opportunity 

 for a mathematical discussion of the theory 

 of chances, for there may, of course, be a 

 considerable number of accidental agreements 

 in two sets of directions such as are here con- 

 sidered. On the third point, it is only reason- 

 able to ask that joints should be shown com- 

 petent to determine the direction of the drain- 

 age lines under consideration, for if they 

 possess no such competence, then all the coin- 

 cidences of direction would have to be at- 

 tributed to chance. The problem here appears 

 to be briefly as follows : Given an ancient land 

 area, subject to normal processes of erosion in 

 many successive (incomplete) cycles of erosion ; 

 in what way can the originally consequent 

 drainage be influenced by underground joints ? 

 It can hardly be supposed that the joints were 

 at the beginning of such power as to have 

 determined the course of the original drain- 

 age — that would be possible only if they were 

 gaping fissures or faults. The original drain- 

 age was presumably consequent on the original 

 slopes and inequalities of the land surface. 

 As to streams of later development, it is pos- 

 sibly the case that in a district of resistant 

 rocks, rapidly given great altitude, the head- 

 ward growth of subsequent streams along 

 joints in the walls of the consequent stream 

 valleys might in time replace the minor con- 

 sequents ; but we have at present no suificient 

 inductive basis for such a conclusion; perhaps 

 the case brought forward by Iddings may be 

 of this kind, but there the argument of coin- 

 cidence has not been supplemented by any 

 argument of competence. On the other hand, 

 a region of relatively weak rocks, slowly ele- 

 vated to a moderate altitude — and this is a sup- 

 position appropriate to the Wisconsin district 

 under consideration — the opportunity for the 

 growth of subsequent streams along joint lines 

 would be very poor; the valley sides would 

 usually be cloaked with waste, so that the 

 capacity of the joints to guide the develop- 



ment of new streams would be small as com- 

 pared with the capacity of the existing streams 

 to persist in their courses. There remains 

 the possible tendency of streams to change 

 their courses by lateral shifting so as to come 

 into relation with underground joint systems; 

 but onr this point we have little information. 

 Finally there is the possibility of other causes 

 than joints for stream directions in the dis- 

 trict studied. Some other causes certainly 

 have operated in some instances, for there are 

 various streams indicated on Harder's maps, 

 unrelated to neighboring joint directions (for 

 example. Underwoods creek, neai* Avoca, PI. 

 III.). In this connection there should be a 

 critical examination of the stage of develop- 

 ment of each stream; for while a stream in a 

 vertical-walled gorge might well be closely 

 related to a joint system, it is quite otherwise 

 with a stream that is swinging on the flood 

 plain of an incised meandering valley. 

 Streams of the latter class appear to occur not 

 infrequently in the district studied, but no 

 special mention is made of them in the text : 

 surely the coincidence of an intermeander 

 tangent with a joint direction can mean little ; 

 for the tangent has in all probability changed 

 its direction and its location during incision, 

 and these changes have been almost inevitably 

 controlled by the changes in the adjacent 

 meanders rather than by underground joints. 

 In all these respects additional study is 

 needed before demonstration can be reached. 

 The fact of frequent coincidences, if well at- 

 tested, is therefore a good first step in an 

 inquiry which we hope Harder may continue 

 to follow. 



BLOCK MOUNTAINS IN NEW ZEALAND 



The south-central part of the southern 

 island of New Zealand possesses a group of 

 block mountains, "probably not less remark- 

 able than the ranges of the Great Basin be- 

 tween the Sierra ITevada and the Wasatch," 

 of which a brief description is given by J. 

 Parks, professor of mining in the University 

 of Otago (" The Geology of the Area covered 

 by the Alexandra Sheet, Central Otago Divi- 

 sion," N. Z. Geol. Survey, Bull. 2, E". S., 



