A PATHOGENIC LUMINKSCKNT UACTKRITM. IQ9 



almost every one of twenty separate sand fleas tried, luminous 

 bacteria could be isolated and grew well and produced light on 

 artificial media. The luminous bacteria isolated were quite 

 surely Bacterium Giardi, but more proof is needed to demonstrate 

 that this is the only species of luminous bacteria which inhabits 

 the sand flea. This demonstrates, as pointed out by Dahlgren 

 (3), that these bacteria are not necessarily luminous continuously, 

 although it does not preclude the fact that they may be kept 

 continually luminous without being passed through the sand flea, 

 as has now been done for two years by the author. This obser- 

 vation also seems to point to the conclusion that this bacterium 

 is a normal inhabitant of the body of the sand flea but that at 

 certain times and under certain conditions it invades the muscles 

 and increases in number so rapidly that the sand flea becomes 

 luminous and dies. It is possible that the bacteria do not 

 become luminous until they have spread to the muscular tissue 

 of the sand flea and thus luminosity would mean that the host 

 was doomed to die, since I have never observed a luminous sand 

 flea recover. As to what causes the bacteria to invade the 

 muscular tissue, especially in August at Woods Hole, Massa- 

 chusetts, nothing yet is known. It may be a lowered resistance 

 of the sand flea or increased reproductive activity of the bacterium 

 due to a higher temperature or due to other causes. It should 

 also be said that out of examination of possibly twenty thousand 

 sand fleas only a very few ever become luminescent and it is 

 quite possible that this bacterium plays a part in the death of 

 thousands of sand fleas that never become luminescent. In 

 other words, if this bacterium is really pathogenic, of which 

 there is still some doubt, it is doubtful if there is any connection 

 between luminosity and pathogenicity other than that great 

 numbers of this bacterium under favorable environmental con- 

 ditions naturally produce a maximum deleterious effect upon 

 sand fleas and also give off more light. 



SUMMARY. 



Amphipod Crustacea are the host of Bacterium Giardi which 

 becomes luminous under certain conditions and may kill the 

 sand flea. 



