Peale.] 288 [December. 



His constitution was not robust ; he suffered much from ill health 

 at various periods of his life, the last years of which, were years of 

 suffering, more or less, but it never changed his kindly feelings to- 

 wards his friends, family, or the dumb pets of his household, espe- 

 cially did it never relax his energy in the great and complex business 

 of his firm, or the more important charities of his heart and purse. 



Too much praise cannot be bestowed on the final disposition of his 

 worldly affairs ; he did all that he intended to do in charity and be- 

 nevolence during life, and left to his heirs a bright example for 

 their guidance. In this respect so different from the ostentatious 

 wording of those testamentary dispositions, that are now wasting their 

 ample provisions in political jugglery, or turned aside by mismanage- 

 ment to some other object, perhaps more objectionable. 



His large property has been bequeathed, after provision for two 

 surviving sisters, to and among his widow, two daughters, and an 

 adopted daughter, in the latter, exemplifying that sentiment of jus- 

 tice and affection, that was so characteristic of him all through life. 



Prof. Kendall read the following communication from Mr. 

 George Davidson, of the United States Coast Survey, viz. : 



A New Level for the Measureinent of the Inclination of the Transit 

 Axis of Portahle Transit Instruments, for all Altitudes beyond 

 the Reach of the Ordinary Striding Level. 



The anomalous condition has heretofore existed that all stars best 

 situated for obtaining time from meridian observations (except in 

 extreme high latitudes) were really observed upon with hesitancy 

 and doubt, because the striding level could not measure the inclina- 

 tion of the transit axis, and the inequalities and irregularities of the 

 pivots, when the telescope pointed to a greater altitude than 55°. 

 Between this altitude and the zenith all errors of inclination were 

 necessarily assumed the same. 



In combining zenith stars with quick-moving stars, the errors of 

 level enter largely in giving erroneous values of azimuth and collim- 

 ation : no reliance can be placed upon the value of the coUimation 

 error derived from reversals between sets of zenith stars observed to 

 avoid unknown azimuthal errors; and the irregularity of the value 

 of the collimation error derived from reversals upon different circum- 

 polar stars, beyond the reach of the level, must frequently be attrib- 



