Historical Note on the Method of Least Square 



the illuminatiTig power, compared with olefiant gas, is one-third 

 greater, or as 1 to 1-J, according to Mr. Lewis Thompson's rule. 

 The sp. gr. of hydrocarbons, found by an indirect method, was 

 1-271. The candle-power of the Fort Wayne gas would be, 

 therefore, according to the chemical standard 74-41 candles, or 

 6 to 6 times that of ordinary coal gas. The actual candle-power 

 by photometric test is much higher. 

 The analyses were all made (with one exception) with Prof. 



•d contents of measurer, but a ver;^ small quantity of gas 

 Id be analyzed, — about \ c. c. when nitrogen was present To 

 e an idea of the experimental error on so small a quantity, 

 r analyses of air, by electrolytic H are added. 



Air erap. (expanded vol.) 62'5 Air erap. 



After H, 125-9 After H, 



After explosion, 



87-3 



After explosion, 10-182 



iC=12-87 perct. 



0=20-59 



iC= 1-445 7 perct. 0=21-023 



3d 





4th. 



A..™p,„,ea. 



30-088 



7-547 

 14-790 



After explosion, 



21-028 



10-496 



iC=3-020 pel-ct. 



0=20-368 



iC=l-431 per ct. 0=18-969 



Fort Wayne, Indiana, 



March 20, 1871. 



=== 



Art. LIX.— ^ Historical Note on the Method of Least Squares ; 

 by Cleveland Abbe, A.M. 



, It is well known that the " Method of Least Squares "al- 

 though first published iu printed form by LeGendre m IbUb, 

 iQ his "NouvellesMethodes," was first invented by Oauss as 

 early as 1795, and had for years been taught by him m ms 

 lectures to his students at Gottingen. It was, however, some 

 years before the Gaussian method came into general use and 

 especially were English scholars very slow to acquaint tiiem- 

 selves with its merits. I have, therefore, been much mterestea 

 in finding that, in 1808, Professor Robert Adram, at that time 

 in New Brunswick, N. J., published the ^^^etjiod of lea^t 

 Jiiaresinthe "Analyst," having been iiidfp^*^^5%^^J! 

 this invention by the study of a prize problem ottered some 

 months previously in that periodical ^.nfir^^l 



^ As the editor of, and chief contributor to, the Mathemat cal 

 Correspondent, the Analyst, and the Mathematical Diaiy, and 



