122 bulletin: museum of comparative zoology. 



injected into the circulation of dogs, birds, and other animals, causes 

 convulsions and death, even when in small doses. Mr. A. H. Verrill, 

 of my party, on one occasion saw the venom ejected as a fine spray, 

 from the parotid glands of a large toad, when it was much irritated." 



For a good account of the histology of these poison glands, and a 

 fairly full bibliography on the subject, see Shipley and Wi.slocki (: 15). 



The eggs are laid in water in the spring. Standing fresh water is 

 scarce in Bermuda, but there are places where pools exist in rainy 

 weather. This water is not always fresh, but brackish water is not 

 refused, and it is said that toads will sometimes lay their eggs in the 

 salt water of mangrove swamps. 



The development of an amphibian in salt water seems very unusual, 

 l)ut Glaser (: 12) discusses a case of development of frog larvae, de- 

 scribed by Pearse (: 11), in the mouth of a creek emptying into Manila 

 Hay Avhere the water was partly salt. To account for this Glaser 

 experimented on the larvae of Ratia pipirns and draws the following 

 conclusions : 



" The ability of amphibian eggs to develop in sea water is dependent 

 on the principle of ionic antagonism. In addition to this, however, 

 their power of acclimatization plays an important role, for it not only 

 enables them to withstand the passage from dilute to strong solutions, 

 but the opposite process as well If in addition to this we re- 

 member that the species found by Pearse is probably racially accli- 

 mated to the conditions under which it lives, his findings do not appear 

 inexplicable." 



The breeding season of Bvfo agua depends largely upon atmospheric 

 conditions. If there is plenty of rain it may begin in February, but 

 if there is continued drouth it may be delayed until July. April is 

 about the usual time. 



Clark (: 16) and Ruthven (: 16) have collected data that show the 

 \ariability of the breeding season. 



(31ark quotes a letter from Barbados as saying: 



"It certainly lays its eggs at least twice a year. In August last 

 year (1914) ponds in Barbados were full of tadpoles, and again in 

 February this year (1915) the same thing was noticed." 



In Trinidad the breeding season is reported as August-October. 

 In Demerara " it appears to breed about the commencement of the 

 wet season, .somewhere in November or thereabouts." 



Ruthven has made some observations on the habits of the species 

 in Demerara: 



"On the Demerara River, about thirty-five miles south of George- 



