256 ANNUAL REPOET SMITHSONIAN INSTITUTION, 1932 



was a great advance when the use of unprotected thermometers be- 

 came general a few years ago. This method really measures the 

 pressure existing at each instrument of a given series and from these 

 the depths are calculated. The pressure of the water compresses the 

 glass of the unprotected thermometer and makes it read higher than 

 the i^rotected thermometer with which it is paired. The difference 

 in reading between the two instruments can then be translated into 

 depth and with a surprising degree of accuracy. Not only are re- 

 liable depths now secured for the temperature observations taken 

 from a modern research ship, but of course these same depths are 

 good for the samples brought up by the water bottles. The water 

 samples are of sufficient volume so that besides salinity, several other 

 chemical factors can be determined. In fact, all physical observa- 

 tions made with this apparatus are now on a comparatively accurate 

 basis. 



The primary collecting instrument of the marine biologist is, of 

 course, a net of one sort or another. Many different types are em- 

 plo3''ed, depending on what type of animals or plants are sought and 

 on Avhether the collections are to be obtained from the bottom, the 

 surface, or intermediate depths. Of late years, attention (formerly 

 directed chiefly to the bottom fauna) has been largely centered on 

 the drifting community of animals and plants — the so-called 

 " plankton " — which are captured as a rule with some form of 

 tow net. Tows made deep down in the water have, until recently, 

 been confused by the same inaccuracy as have the routine physical 

 observations, namely, the uncertainty as to the precise depths from 

 which the captured specimens came. Not only was the maximum 

 depth reached by the net uncertain but also the animals might have 

 gotten in on the way down or on the way up. This last source of 

 error is especially serious in the case of deep-sea tows, for with these 

 the time involved in lowering and raising the net may be several 

 times as long as the horizontal tow itself. Even for plankton work 

 near the surface, it has been difficult to know at just what depth 

 the net had been fishing or to do accurate quantitative work with 

 more than one net at a time. Improvement in the technique of 

 tow-net work has come gradually and is perhaps not yet at a satis- 

 factory stage of development. 



One of the first steps was the perfecting of nets which can be 

 sent down closed, opened during the tow, and closed again befoi-e 

 being hauled to the surface. Another improvement has been the 

 use of a kind of bucket at the back end of the net wliich protects 

 the catch from being rubbed too much against the netting during 

 the long haul to the surface. As many deep-sea animals are ruined 

 as specimens through this cause as through the release in pressure 



