MORTALITY OF FISHES, WEST COAST OF FLORIDA. 17 



a disease would confine itself to a limited locality for the larger 

 part of a century or perhaps a much longer period of unrecorded 

 observation. 



DILUTION OF THE WATER. 



4. Dilution of the water hardly deserves serious consideration, for 

 the phenomenon is not correlated with the rainy season or unusual 

 discharge of the rivers. As shown by the table, the salinity does not 

 indicate serious dilution of the water. Of course a dilution ade- 

 quate to explain the mortality of fishes would concern only an excess 

 of run-off above the normal. The area of ocean concerned, in com- 

 parison with the small rivers, at once negatives this hypothesis : and 

 if the water were sufficiently diluted, it may be assumed with good 

 reason that unconfined fishes would seek their proper salinity by 

 migration. 



VOLCANIC AND NATURAL GASES. 



5. It is now time to call particular attention to certain circumstances 

 of prime importance that have served largely to destroy the fore- 

 going suggested causes, and which can not be overlooked in arriving 

 at a conclusion. They are (a) the irregular periodicity, both as to 

 years and seasons; (h) the strictly marine aspect of the phenomenon; 

 (c) the large area covered; (d) the definite limitation of this local- 

 ity; {e) the limitation of the mortality to the animals, whose res- 

 piration is performed by an oxygen carrier, haemoglobin, ha?mocyanin, 

 etc. ; (/) the progressive southward appearance of the mortality. 



The irregular periodicity, the marine aspect, the area covered, and 

 particularly the limitation to the region concerned all suggest a geo- 

 logical explanation, whether the issue of a gas or the occurrence of 

 an earthquake, fixed in one place, sporadic in outbreak, and inde- 

 pendent of weather fluctuations. The species affected and the south- 

 ward progress are not contradictory to such an explanation, as will 

 be seen. 



Dr. T. W. Vaughan, of the United States Geological Survey, stated 

 verbally to the writer that he saw only the remotest possibility of 

 a volcanic gas in this region. The region is inactive volcanically, 

 and the discovery of a volcanic fumarole would occasion great sur- 

 prize to geologists. Beyond that we have nothing for or against 

 volcanoes. There is no question that volcanic gases would be deadly 

 to fishes. For, while volcanoes emit different gases in different stages 

 of their activity, the emissions always contain substances deadly to 

 fishes, if nothing worse than carbon dioxide It is scarcely necessary 

 to present data here as to the composition of such gases. Any chlo- 



