162 PHYSICAL INVESTIGATIONS. 



an hydrometer of the ''Challenger "* type^ kindly presented to the 

 Association by J. Y. Buchanan, Esq., F.R.S. This instrument 

 weighs in vacuo 150'6897 grammes, and is furnished with seven 

 brass weights varying from 0-3102 gr. to 4-0100 gr., giving for each 

 sample of water at any temperature at least two distinct determina- 

 tions. The constants of the instrument have been carefully determined 

 twice at temperatures varying from 5° C. to 30° C, and the maximum 

 probable error has been found to be 0-00005, a result agreeing with 

 that arrived at by Mill.f 



The want of a sufficiently delicate and reliable balance made the 

 further examination of samples at first almost impossible. Through 

 the kindness of Messrs. Balkwill, chemists, Plymouth, we now have 

 access to an instrument giving results reliable to O'l mgrm. ; and 

 recently Prof. A. M. Worthington, of the Naval Engineering College, 

 Devonport, has lent to the Laboratory a balance capable of weighing 

 100 gr. to within 1*0 mgrm. While we are unable to carry out 

 analyses with the high precision reached, for example, in Dr. Gibson's 

 work for the Scottish Fishery Board, we may now hope to make 

 determinations sufficiently accurate to be of considerable value. The 

 first object aimed at is to obtain determinations of density by means of 

 the modified form of SprengePs pyknometer.J In the hydrometer 

 determinations given, the densities are i^educed to 15-56 C. com- 

 pared with distilled water at its maximum density point by Dittmar's 

 tables. § These tables, however, can only be safely used where the 

 sample approaches the standard density of 1-02600 at 15-56° C. In 

 all other cases determinations made with Sprengel tubes filled in 

 melting ice are much to be preferred. 



In the more strictly chemical work I have had the advantage of 

 the co-operation of Mr. F. Hughes, the Chemist of the Association. 

 The alkalinities of all the samples already collected have been 

 determined by the usual methods with a fair degree of accuracy, 

 and considerable progress has been made with estimation of the 

 amounts of chlorine. I have thought it best to defer publication of 

 the chlorine results until more progress has been made with the 

 densities. 



Table I gives the results of observations made in a trip on 

 ss. " Deer-hound '^ in the Channel in June last. On June 15th a 

 line was taken from Bolt Head to a point west of Hanois Light, 

 and thence to St. Peter's Port, Gruernsey ; on June 16th from 

 Guernsey to St. Catherine's in the Isle of Wight; and on June 17th 



* Challenger Reports, Narrative, vol. ii, pt. 2. 



t Proc. Roy. Soe. Edin., xiii, p. 35. 



+ Report of the Fishery Board for Scotland, 1887, p. 336. 



§ Challenger Reports, Physics and Chemistry, vol. i, p. 70. 



