440 The Philippine Journal of Science i»is 



various leaks was thus partially removed. In spite of the daily 

 circulation established, it was possible to obtain samples of bilge 

 water containing sufficient decomposition products for partial 

 identification. It was found that bacterial action was taking 

 place throughout the mass of copra, with the production of a 

 large amount of hydrogen sulphide. Distinct tests for this gas 

 were obtained in the open air at a distance of about 150 meters 

 from the scene of the wreck, and vessels stationed near to render 

 assistance were completely blackened within a short time. The 

 hydrogen-sulphide-producing organism was found to be a motile 

 rod. It appeared to act on the cellular tissue and cause an ex- 

 traordinary selective destruction of the copra. One of the crew 

 volunteered to obtain a sample of bilge water early in the morning 

 before pumping was started. Accordingly, I left a bottle with 

 him the preceding evening, with directions to fill it and im- 

 mediately wire on the rubber stopper. This sample was car- 

 ried to Manila for experimental purposes, and opened some days 

 later in the laboratory. It consisted entirely of coconut oil, 

 of slightly dark color and disagreeable odor. No water was 

 gathered with the oil, indicating that the amount of free oil in 

 the hold at that time must have been considerable, but owing 

 to the darkness its presence had not been noted. Unfortunately, 

 the sample bottle was made of opaque blue glass, with the result 

 that I had no intimation of its true contents until after my return 

 to Manila. Moreover, the floating debris of discharged copra 

 had already rendered the adjacent surface of the ocean oily and 

 effectively masked the effect of free oil which must have been 

 pumped out early in the morning. The bacterial action had 

 broken down the cellular tissue of the copra and liberated the 

 oil. This oil contained 60 per cent free oleic acid. 



The water passing through the copra was decidedly acid, a 

 specimen taken late in the afternoon containing 0.5 per cent 

 calculated as sulphuric acid. The acidity must have been much 

 greater before the pumps were installed, and accounted for the 

 caustic action of the water. Sulphurous and sulphuric acids were 

 identified in the water, accompanied by various organic acids with 

 indol and skatol resulting from decomposition. The action of this 

 poisonous water was very severe, causing intense smarting where 

 it came in contact with the skin, accompanied by very pronounced 

 swelling of the testicles. These evil effects were mitigated by 

 frequent bathing in salt water, and a shower was maintained on 

 the deck for this purpose. This acid water was being pumped 

 continually into the open sea, and the decaying copra was dis- 



