EXPERIMENTS IN THE KEEPING OF PLANKTON ANIMALS. 563 



tures of Bacteria raised on peptone. But the results of these experiments 

 showed that species were capable of tolerating such conditions to a much 

 greater degree than was expected, and did not bear out the supposition 

 that bacterial infection had been the primary cause of failure. Experi- 

 ments also made with and without an air-supply gave no definite evidence 

 of the value of an air-supply as improving the conditions in water that 

 was naturally aerated at the outset. 



There remained a possibility that harmful conditions might be pro- 

 duced through the accumulation of excretory products in the vessels, 

 and to test this a special experiment in Position A was arranged, by 

 which a constantly chariging supply of water and food was passed 

 through the vessel in wliich the animals were contained. For this pur- 

 pose a large bottle, of 10 litres in capacity, was filled with sterilised 

 " outside " water, with Miquel- Allen solution in proportions one-third 

 of those used by Allen and Nelson, and this water was infected with a 

 culture of Nitzschia. The vessel of the experiment, in which 5 Calanus 

 were placed, was a 1-5-litre flask, and into this the supply was led by a 

 glass tube from the large bottle, the apparatus being arranged in the 

 form of a self -regulating siphon, by which the water in the flask was 

 kept at a constant level. From close against the bottom of the flask, 

 a siphon tube was led out from the flask to the exterior, to serve as a 

 waste pipe, and this, drawing on the contents of the flask, was so ad- 

 justed by a screw clip that the water and food-supply was made to pass 

 through the flask at the rate of 1 litre per diem, nearly. An air-supply 

 was also provided at a slow rate in the flask by connection with the air- 

 pressure system of the Laboratory. A good light was ensured and, the 

 food growth being strong, the conditions were such as appeared to 

 meet afl requirements. But little improvement was shown in the result 

 of this experiment which, though more successful than most of the 

 previous ones, produced only a maximum life of 28 days, and an average 

 of about 21 days, the shortest life being 14 days. This experiment was 

 started on October 16th, 5 days prior to the commencement of an ex- 

 periment made in the same position, in a 1-litre flask under the ordinary 

 conditions, in which 3 Calanus subsequently showed a maximum of 84 

 days, and an average of 44 days (p. 560). 



Position C. 



It had in the meantime been observed that a Calanus which had 

 been left in a 2000 c.c. jar, partly submerged in one of the tanks in the 

 General Laboratorv on August 28th, was still alive on October 18th, 



