368 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.xxvii. 



That article relates to the larval and pupal stages and the conditions 

 under which they live. These facts will not be restated here, as we 

 have nothing new to add, but there are certain other aspects of the 

 habits of fleas which very much need a fuller discussion. 



In recent discussions of the relations of parasitic and other insects 

 to the transmission of disease, much space has been given to flies, 

 especially mos({uitoes, and but little to fleas. While as a matter of 

 fact the latter ma}^ be of even greater importance, not onh' because of 

 their more insidious attacks, but also because of their association with 

 some of the most terrible diseases. 



The whole matter rests upon the host relations of the various species. 

 It is a well known fact that the cat and dog flea will take very readily 

 to the human being. It is to be noted that the cat and dog flea is 

 closely related to Pulex irritans and similar to it in the more impor- 

 tant details of structure. Some of the rabbit fleas, which are likewise 

 closely related to P. irritans^ will also readily attack the human being, 

 which I had occasion to learn as a youth. While these fleas will 

 remain on a human being for some little time and bite frequently while 

 there, still they do not habitually frequent that host and his clothing 

 and bed as does P. irrttanst. 



Among closely related animals or animals of very similar habits one 

 species of flea may have a mimher of normal hosts. But our knowl- 

 edge of this matter is exceedingly fragmentarv and uncertain. Cases 

 of temporary' hosts being as common as the}- are, it becomes very 

 probable that many of our records refer merely to the temporary host. 

 A rabbit running into a badger hole, a mouse into a mole burrow, an 

 owl eating a mouse, a cat devouring a rat — these andman}- other fortu- 

 itous circumstances furnish conditions favorable for at least temporary 

 transference. I have referred to a number of these cases in the 

 account of the separate species. 



The character of the hair and thickness of skin was at first consid- 

 ered as controlling the range of parasites, due to the close relation 

 these conditions nmst have to the structure of the flea, especially the 

 length of mouth parts and covering of bristles. Exceptions were soon 

 found to this rule, though in general such relations may be said to exist. 



Excluding the accidental records on carnivorous animals, we may 

 say that in the United States the cat, dog, and rabbit fleas are closely 

 related, to P. irritans anA will readil}' attack the human being, while 

 the mouse, rat, squirrel, mole, and shrew fleas are not closely related 

 to P. /'rritan.<i and have never been known to })ite the human l>eing. 

 Mr. F. H. Chittenden wrote me that he believed rat fleas sometimes 

 bit human beings in Washington, but he has not yet verified this state- 

 ment by the actual capture of specimens in the act. Nor do I know 

 of any such records. 



South of the United States conditions are wholly different, and 

 present an aspect of considerable economi'^ importance, for here we 



