220 GENERAL CONDITIONS OF SCIENTIFIC RESEARCH. 



that time there were not more than ten persons who could 

 fully understand it, and it was fifty years before it was 

 fully appreciated. Hunter's experiments, &c., also, in 

 comparative anatomy, cost him 70,000. in money alone, 

 besides an immense amount of personal labour. It was 

 by making at least one thousand observations, that 

 Graham, in the year 1722, discovered the diurnal variation 

 of the magnetic needle ; and Canton, about the year 1756, 

 by making at least four thousand observations, confirmed 

 that discovery, and also discovered the yearly variation of 

 the needle, and that the diurnal variation was greater in 

 summer than in winter. ' The labour of comparing with 

 theory above eight thousand observed places of the moon, 

 by computation of the same number of places which had 

 been observed at Greenwich, between the years 1750 and 

 1830, occupied a number of calculators (sometimes as 

 many as sixteen) during about eight years, and was made 

 for the purpose of correcting the Lunar Tables. One 

 result of these calculations was the discovery of two lunar 

 inequalities, both due to the attraction of Venus.' Several 

 years have already been occupied by a number of calcu- 

 lattfrs in computing the results of the observations of the 

 late transit of Venus, and the labours are not yet com- 

 pleted. The discovery of thallium and the accurate deter- 

 mination of its atomic weight cost Crookes many years of 

 labour and a large sum of money ; and the determination 

 of the chief properties of anhydrous hydrofluoric acid, 

 fluoride of silver, and other fluorides, occupied the author 

 of these pages most of his time during nine years, and 

 necessitated the purchase of several hundred ounces of 

 platinum vessels. The compilation and publication of the 

 mere list of the titles of original scientific investigations 

 made since the year 1800, together with the names of the 

 investigators, and entitled ' The Royal Society Catalogue 



