METHODS AND DATA. 



3 



The total number of wliat I term " comparable hauls " with tow-nets is made up 

 as follows : — 



25 hauls at fatlioms~l 



I 



I Epiplanktou hauls 

 ' I with open tow-nets. 



! 

 > J 



various depths, open serial tow-nets. 



„ closing Mesoplankton net. 



127 



These omit as " non-comparable " from the list in Table II., two hauls of which only 

 part was kept (21 d, 21 cV), two miss-fires of the Mesoplankton net (27 «, 27 a), three 

 hauls of the trawl (29 a, 30 a, 30 e), and four hauls of which the specimens cannot at the 

 moment be traced {cf. p. 11). 



As regards the temperatures, we depended on the daily observations in the ship's 

 log for the surface records. But as a Quartermaster's readings of a thermometer are not 

 invariably accurate, I have generally taken a mean for each day as the safest approxi- 

 mation, and as sufficiently useful for zoological, although not for meteorological, work. 



As regards the temperatiires at 25, 50, and 75 fathoms, I regret to say that in the 

 constant calls on my time in connection with the rest of the work, I entirely forgot to 

 take any observations ! The temperatures for these depths have, liowever, beea 

 supplied as mean values for July in this district by the kindness of Dr. Schott, of the 

 Deutsche Seevvarte in Hamburg. 



The deep-sea temperatures are taken from a very complete series of observations which 

 the officers of H.M.S. ' Research ' were kind enough to make for me at the time. The 

 thermometers used were of the Miller-Casella type, and included some special ones witli 

 a scale of twice the usual size. They were used on sounding-wire ; attemjits to use the 

 -^i' diarn. wire rope showed tliat the vibration of the rope jarred the indices down and 

 rendered observations quite unreliable. 



The actual observations were as follows, omitting those on the wire rope : — 





Depths in Fathoms : Temperatures in Degrees Fahrenheit. 



These observations have been plotted into the curve marked I. in PI. 1. Alongside of 

 this has been plotted a curve, marked II., deduced from the much scantier observations 



