39 2 WILL SCOTT. 



convection of a closed room, i. e., a lower and an upper current 

 moving in opposite directions. When the air currents were too 

 weak to be measured, their direction was observed by means of 

 a candle flame. 



These results confirm and elaborate the results of Banta ('07) 

 and Egli ('04). Banta found by weekly observations in May- 

 field's Cave that the air currents reversed about October I and 

 April i. Egli found in Holl-Loch, a cavern explored by him, 

 that the currents were down when the outer temperature was 

 8.1 Centigrade, and up when 2.3 Centigrade or below, being 

 very strong at --8 Centigrade. 



Method. - - It is much more difficult to collect plankton in a 

 cave than in ordinary waters. The lack of light, the dimensions 

 of pools, the relatively small amount of plankton, the great and 

 sudden variations in the cave stream, and the necessity of trans- 

 porting and manipulating the apparatus unassisted, rendered many 

 of the ordinary methods impracticable. 



After various experiments the method finally adopted for all 

 quantitative work was as follows : A net was constructed of bolt- 

 ing silk No. 20 (Dufour) after the pattern described by Kofoid 

 ('97). The net was suspended in a pool with about 3 inches of 

 its filtering surface exposed. Two hundred gallons of water were 

 dipped and poured into it with a bucket (i y 2 gal. capacity) at the 

 rate of 10 gals, per minute. The catch was preserved in a 

 4 per cent, solution of formalin. I am well aware that this method 

 results in some error (Kofoid, '97), but it was the most practicable 

 for this investigation. 



For the examination of material, glass troughs were constructed 

 by cementing glass strips on a slide, with marine glue. The 

 width approximated the diameter of the field of the microscope 

 (2^ obj., 2 in. ocu.), the length 40 mm., and the depth about I 

 mm. The material was pipetted into this trough until the sur- 

 face film was parallel to the slide. The entire catch was counted 

 in quantitative work. 



When higher powers of the microscope were necessary for 

 careful study, the material was removed from this trough to a 

 slide on which strips of cover glass were glued for the support 

 of the cover glass. After examination, the material was washed 

 from the trough into a new vial with 4 per cent, formalin. 



