234 BULLETIN OP THE NATURAL HISTORY SOCIETY. 



division of the province into districts. In a work now nearing com- 

 pletion I have had to make such a division, and thus have given some 

 study to the subject, with the following results. For a detailed 

 natural division the counties, with their artificial boundaries, are 

 not available; and the geology, with its correlated topography, is too 

 irregular and complicated. We do find, however, a convenient and 

 fairly natural division in the river systems, which for historical pur- 

 poses is especially suitable, since they have powerfully influenced the 

 distribution of settlements. But if we attempt to separate these 

 systems by lines drawn along their watersheds (as is done on the 

 accompanying map, fig. 1), we see at once that in a general way there 

 is a correspondence between counties and river systems. This is, of 

 course, no mere accidental coincidence, but is the result of a true 

 causal relation, for the principle that has determined the setting off 

 of our counties has been that of grouping them around the river basins 

 and running their boundaries along the watersheds.* It will obvi- 

 ously be convenient in making our natural divisions to pay as much 

 attention as possible to the familiar county divisions, and hence the 

 boundaries of the latter may well be used to settle doubtful points in 

 the natural divisions. There are two cases in which it is profitable to 

 use the county divisions to determine details of the natural divisions 

 —first, in the points of separation of the systems along the sea coast 

 (for here the county lines have been very wisely chosen), and second, 

 in the secondary division of the large St. John system. The geology 

 also, in doubtful cases, may be called to aid. Taking into account all 

 of these factors, the river system districts of New Brunswick would 

 appear to be as follows, and as traced on the accompanying map. 

 Natural divisions should obviously be known by indigenous names, 

 for which, happily, convenient Indian names are available. 



I. The Passamaquoddy (or Charlotte) District. Includes the basins of all 

 rivers from the Oobscook to Point Lepreau. 

 77. The Woolastook (or St. John) Dint rid. Includes the entire St. John 

 System and the smaller rivers from Point Lepreau to Martins Head. 



* That river systems and counties do not correspond yet more closely is due to three 

 causes :— First, county lines are, for convenience of running and marking, best made 

 straight, while watersheds are crooked ; second, the geography of ihe province was im- 

 perfectly known wheu some of the lines were established by law, and hence they do not 

 run us it was supposed they would ; third, some of our rivers run so far across the prov- 

 ince and head so near the basins of others, it is practically more convenient to allow their 

 heads to fail into other counties. Such is the case with the St. ('mix and Magaguadavic, 

 Washademoak and Salmon River, Miramichi, and Restigouche. 



