bromine, or iodine), and phosphorus by the 

 procedures of Vogel (1956). Positive tests were 

 obtained only for phosphorus. In all cases, the 

 results were compared with known compounds 

 to verify the test. 



ELEMENTAL QUANTITATIVE 

 ANALYSIS 



Semimicro quantitative analyses for carbon, 

 hydrogen, and nitrogen were performed by 

 Peninsular Chem Research, Inc., Gainesville, 

 Fia. (table 2). 



Organic phosphorus was determined by a 

 conventional method (Martin, 1968). The cali- 

 bration curve (fig. 1) was prepared with .a 



10 20 30 40 



PHOSPHORUS UGi IN SO-Ml STANDARD SAMPLE 



Figure 1. — Calibration plot for organic phosphorus 

 determination. Standard sample prepared by decom- 

 position of 9.8 mg. of disodium p-nitrophenylphos- 

 phate, followed by dilution to 50 ml. with triple- 

 distilled water at 705 m/i. 



phosphorus standard (disodium p-nitrophenyl- 

 phosphate). Exactly 0.89 mg. of substance II 

 was decomposed and diluted to 50 ml. ; two 

 25-ml. portions had absorbances of 0.058 and 

 0.055, which corresponded to 1.41 percent P. A 

 second analysis was performed similarly with 

 0.99 mg. of substance II, except that 6.968 /.tg. 

 of phosphorus (as phosphorus standard) was 

 added; total absorbances of 0.115 and 0.118 



Table 2. — Elemental quantitative analysis of substance 

 II (bloom sample) 



Hydrogei 



6.94 

 7.60 

 0.00 



1 Not analytically significant. 



2 Calculated by difference from 100 percent total. 



436 



were obtained, which gave a corrected value 

 of 1.35 percent P. 



Total oxygen was determined by Schwarzkopf 

 Microanalytical Laboratory, Woodside, N.Y. 

 Exactly 0.200 mg. of toxin sample produced 

 0.115 mg'. of carbon dioxide. 



MOLECULAR WEIGHT 



The molecular weight of substance II (bloom 

 sample) , determined by Galbraith Laboratories, 

 Inc., Knoxville, Tenn., by vapor-pressure os- 

 mometry, was 650 (chloroform solvent, 0.03 M) . 



ABSORPTION SPECTRA 



The carbon tetrachloride solution of sub- 

 stance II from the bloom sample had no absorp- 

 tion in the visible region of the spectrum; its 

 absorption maximum was at 269.5 to 270 ni/i 

 (2.5 mg./3 ml. in CCI4) in the ultraviolet 

 region. When the solution was diluted, the ab- 

 sorption maximum broadened and was 265 to 

 270 m/i for the most dilute solution (table 3) . 



The prominent absorption bands of the in- 



Table 3. — Ultraviolet absorption of carbon tetrachloride 

 solutions of substance II at 269 nin 



frared absorption spectra of various toxin frac- 

 tions (as carbon tetrachloride solutions) were 

 determined (table 4). Tentative assignments 

 are also indicated and are compared with 

 standard values given by Bellamy (1958). 



The similarity of substance II samples iso- 

 lated from bloom water and from culture 

 medium is indicated by the infrared data ob- 

 tained with a Perkin-Elmer 337 infracord. Sub- 

 stance I could not be isolated in sufficient quan- 

 tity from culture medium for analysis. The 

 bloom sample of substance I seems to be dif- 

 ferent from the bloom sample of substance II; 

 the infrared spectrum of the former indicates 

 the presence of a carbonyl group, whereas the 

 spectrum of the latter does not. 



U.S. FISH AND WILDLIFE SERVICE 



