234 BOTANICAL GAZETTE [September 



an intermediate step in photosynthesis. If the latter, it should 

 not appear if the experiments were carried out in the dark. To 

 test this, boxes were constructed which were light-tight, but which 

 would allow a ready flow of water through them. These boxes 

 were i ft. square and 18 ft. long. The ends were closed by light 

 traps, the baffle boards of which were inclined in the direction of 

 flow of the water (fig. 3). The lids were also light-tight. All 



Direction of Flow 



Fig. 3 



holes and cracks were closed with pitch, and the whole interior 

 painted a dead black. The boxes were weighted so as to just float ; 

 the waves washed entirely over them except when the water was 

 perfectly calm. The boxes were anchored in the bay (Friday 

 Harbor, Washington) where the tidal currents are heavy, so that 

 there was practically always a flow of water through the boxes. 

 The boxes were large enough to hold several specimens without 



* 



impeding the flow of water. 



In the first experiments, the top 18 inches of stipe from 10 kelps 

 were filled with air by displacement with sea water and corked and 

 placed in the dark boxes. Half of them had the fronds 

 After 5 days in the dark they were removed and the gas analyzed. 

 All showed carbon monoxide. The range was from 0.4 to 1.7 

 per cent, with an average of 0.7 per cent. The 20.9 per cent 

 of oxygen in the air with which they had been filled had prac- 

 tically disappeared, and there was about 4 per cent of carbon 

 dioxide. The oxygen was used by respiration and decay processes. 

 These data were checked by extended experiments, and it was 

 made certain that in the dark as well as in the light carbon monoxide 

 was formed regularly in the air-filled section of the stipe, and that 

 there was no relation between its appearance and the presence or 

 absence of the fronds. 



remov 



