NovEiffiER 27, 1908] 



SCIENCE 



757 



the estimated rate being of more than ap- 

 proximate value. In the case of the main 

 cataract, there have been many cycles, so that 

 an error in one is not fatal to the determina- 

 tion of a mean rate. 



In his recent report upon the recession of 

 Niagara Falls,' Dr. Gilbert devotes much 

 space in trying to prove as erroneous Professor 

 Hall's survey (1842), vehich shows a shelf 

 since fallen aveay. "Without discrediting 

 Hall's work, it seems much more probable that 

 a cycle of undercutting had commenced prior 

 to his survey, though the floor of the river fell 

 afterwards, so that his measured amount of 

 recession belongs to a longer period. The pro- 

 truding sheK shown by Hall is also indicated 

 on the map of the International Boundary 

 Survey of 1819. On this basis the mean rate 

 would be reduced to 0.34 foot (or for com- 

 parative purposes to 0.26). It we were to esti- 

 mate the rate of recession of the smaller cata- 

 ract in terms of the greater, it would be 0.28 

 foot a year (or for comparison, 0.22 foot). 

 On the basis of 0.34 foot, the age of the Falls 

 of Niagara would be found to be 37,000 years, 

 which nearly approaches that already stated in 

 my book. If 0.28 be taken, the result would 

 of course be the same as that based on the 

 main cataract. Conclusive measurement could 

 not be made during a lifetime, besides which 

 the spoliation of the cataract for power pur- 

 poses will destroy the natural conditions; yet 

 additional data are obtainable, so that it will 

 bj possible to write a supplementary chapter 

 on the recession of the American Falls, now 

 that the collateral evidence is called for. 



There are other factors in the physics of the 

 falls of differential value, but they are all 

 embraced in the actual work done by the two 

 cataracts, the relative efficiencies of which do 

 not greatly differ, as shown above, upon the 

 problem being analyzed scientifically. This 

 opportunity of confirming my previous calcu- 

 lations affords me great satisfaction. 



As to the recession surveys of Niagara 

 Falls, it may be stated that mine of 1904 was 

 the fifth one of the main cataract ever made 

 (Preliminary Eeport of the Geological Survey 

 of Canada for 1905, published in the summer 



' Bulletin cited before. 



of 1906), preceding by a few months that of 

 Mr. W. Carvell Hall. In using this last- 

 mentioned survey, Gilbert computes his rate 

 of recession from only that portion of the 

 falls where the greatest depth of water occurs, 

 although that portion is some hundreds of feet 

 less than the diameter of the cataract which 

 has been making the gorge of an equal width. 

 In his calculation, he also uses the mean of 

 ordinates unequally placed. Thus he gives a 

 result of 5.3 feet a year, but on page 25 he 

 adds that there is an uncertainty of one foot. 

 This estimate of five feet a year affords no 

 coefficient of recession throughout the gorge. 

 If corrected for the full width, the rate would 

 be found to approximate that of mine. He 

 also says that the recession between 1875 and 

 1905 was greater than in the previous thirty- 

 three years; but had he broken up the former 

 period into two divisions of fifteen years each, 

 he would have found a great reduction in the 

 rate of recession during the latter fifteen 

 years. In his application of the rate of five 

 feet in determining the recession of the Amer- 

 ican Falls the result would give too short a 

 time, had there not been a previous accelera- 

 tion, due to the higher stage of the cataract 

 during its earlier history (unknown until dis- 

 covered by my soundings under the falls). 

 These compensating omissions, however, give 

 an approximately acceptable rate (0.32 foot). 

 Thirdly, my critic says : 



If the efficiency of Niagara in producing reces- 

 sion varies according to the law, as the efficiency 

 of a river in transportation, Spencer's estimate of 

 the age of the river should be multiplied by a 

 factor larger than four. 



Lake Erie is a settling basin, so that only 

 after occasional severe storms does the lake 

 send down its turbid waters into Niagara 

 Eiver, and then only for a short time, but 

 this small quantity of detritus, even then, 

 varies with the volume and velocity of the 

 river and is lost in the general averages of the 

 discharge measurements. The volume of 

 detritus removed from the channel of the 

 river since its birth is less than a thousandth 

 part of the volume of water that flows down 

 each year. Again, it requires 900 tons of 

 water to loosen every pound of rock carried 



