' [75] TJ. S. FISH COMMISSION 8TEAM?:R ALBATROSS. 77 



'frame in tbe baud or swiugiug it rapidly about tbe bead ; aud if tbis 

 fails more vigorous tapping is apt to follow witb various active meas- 

 ures, none of wbicb tend to improve tbe general condition of tbe instru- 

 ment. 



Tbe indices are also liable to move if tbe instrument is subjected to 

 rougb treatment, altbougb tbis is not of frequent occurrence witb care- 

 ful bandling. Most of tbe minor casualties to wbicb tbe instrument is 



liable are apparent to tbe eye and are readily adjusted. 



i ,■ ' 



WATER DENSITIES. 



Hilgard's ocean salinometer (Plate XLI) is used on board of tbe Al- 

 batross for observing tbe density of sea-water. 



An excellent description of tbe apparatus is given by Prof. J. E. Hil- 

 gard in tbe United States Coast Survey Report for 1874, and reproduced 

 in Sigsbee's Deep-sea Sounding and Dredging-, as follows : 



''Tbe density of sea-water in different latitudes and at different deptbs 

 is an element of so great importance in tbe study of ocean pbysics as 

 to bave caused a great deal of attention to be paid lately to its determi- 

 nation. Tbe instruments employed for tbe purpose bave been, almost 

 witbout exception, areometers of various forms. Tbe differences of 

 density as arising from saltness are so small tbat it is necessary to bave 

 a very sensitive instrument. As tbe density of ocean water at tbe tem- 

 perature of 60° Fabr. only varies between tbe limits 1.024 and 1.029, if 

 is necessary, in order to determine differences to tbe bundredtb part, tbat 

 we sbould be able to observe accurately tbe balf of a unit in tbe fourth 

 decimal place. Tbis gives a great extension to tbe scale, and involv^es 

 tbe use of a series of floats if tbe scale starts from fresh water, or else 

 the instrument assumes dimensions wbicb make it untit for use on board 

 ship. Witb a view to the convenient adaptation to practical use this 

 apparatus has been devised for tbe Coast Survey by Assistant Hilgard. 



"The instrument consists of a single float about 9 inches in length. 

 The scale extends from 1.020 to 1.031, in order to give sufficient range 

 for the effect of temperature. Each unit in tbe third place, or thou- 

 sandths of the density of fresh water, is represented by a length of 0.3 

 of an inch, which is subdivided into five parts, admitting of an accurate 

 reading of a unit in the fourth place of decimals by estimation. 



"The float is accompanied by a copper case, witb a thermometer in- 

 serted within the cavity, wbicb is glazed in front. In use tbe case is 

 nearly filled with water, so as to overflow when tbe float is inserted, 

 the reading then being taken witb ease at the top of the liquid. 



"For convenience and security, two such floats and a case are i^acked 

 together in a suitable case, and a supply of floats and thermometers 

 securely packed in sawdust is kept on baud to replace tbe broken ones. 



"The following table has been derived from tbe observations of tbe 



