of forestry and statistical training, which qualifies them better for both classes of 

 work. Forestry training helps in understanding forestry experiments and statistical 

 training is valuable in mensuration work. 



CONSTRUCTION OF TREE VOLUME TABLES 



Calculation of volume tables for 1 9 Ontario tree species are in progress, using 

 an algebraic method carried out on the I.B.M. computer at the University of 

 Toronto. Measurements of 23,500 trees made by this Branch and by the Timber 

 Branch are available for the purpose. 



The method was originally proposed in 1929 by a professor of forestry 

 employed by the Department, but could not be used because of the heavy arith- 

 metic involved and the non-existence of computers. The method received initial 

 trials in a simplified form in 1950 by the Research Branch and also in the United 

 States and Sweden. The final form of weighted regression was developed by a 

 graduate student with assistance from the Department. 



It is the first time this technique has been used anywhere, and trials indicate 

 that a single computer programme will satisfactorily produce volume tables of any 

 of the several types required for each species. It is also indicated that practically 

 all existing tables produced by older methods should be considered obsolete 

 because of errors introduced by the old methods of construction. 



SAMPLING ERROR OF SYSTEMATIC SURVEYS 



The uniform spacing of the cruise lines along which timber estimates are 

 made tends to give an excessive estimate of probable error when the standard 

 methods of statistics are used. The results of a 100% survey made by the United 

 States Forest Service of an area of eight square miles, have been analyzed in 

 detail to reveal the difference between the results of systematic and random 

 sampling and to test a method of computing standard error worked out by Dr. 

 D. B. DeLury, Chairman, Department of Mathematics, University of Toronto. 



REPORTS 



Research Branch Reports Published During the Year Ending March 31, 1965. 



Fisheries 



Chemical Characteristics of Ontario Lakes as Related to Glacial History. R. A. 

 Ryder. Trans. Am. Fish. Soc., Vol. 93, No. 3, July 1964. Dist. December, 1964. 

 The Invasion of the Lower Great Lakes by the White Perch Roccus americanus 

 (Gmelin). W. B. Scott and W. J. Christie. J. Fish. Res. Bd. Canada 20(5), 1963. 

 Dist. July, 1964. 



Chemical Characteristics of Ontario Lakes with Reference to a Method for Esti- 

 mating Fish Production. R. A. Ryder. Section Report (Fisheries) No. 48. Dist. 

 July, 1964. 



On the Feasibility of Introducing Kokanee the Land-Locked Sockeye Salmon 

 Oncorhynchus nerka kennerlyi to the Great Lakes. F. P. Maher. Section Report 

 (Fisheries) No. 55. Dist. Aug. 1965. 



Status of Fisheries Research Projects for the Year 1963. Section Report (Fish- 

 eries) No. 51. Dist. Aug. 1964. 



Fishes of Northern Ontario, North of the Albany River. R. A. Ryder, W. B. Scott 

 and E. J. Crossman, Life Sciences Contr. 60, 1964. Dist. April, 1965. 



254 



