624 DISEASE IN WILD MAMMALS AND BIRDS 
tories, data usually recorded in terms of the individual 
observer's studies and often inadequate to give the com- 
piler all the facts desired. The same remarks apply to 
the percentage incidence of parasitism. Not to prolong 
the academic discussion, suffice it to say that very much 
the same obstacles present in the second avenue — that of 
pathogenicity of the individual species. Our own data 
referring to this second heading will be presented later, 
but after the failure of the first avenue, the second and 
third lose greatly in value. At best, statistics can be 
only suggestive. Unless critically and suspiciously 
interpreted, and with a full appreciation of their limita- 
tions from a foreknowledge of the way in which they were 
compiled, they would only delude the reader and offend 
science, and so we abandon this line of reasoning. 
At present the best results of the study of patho- 
genesis by animal parasites will probably be reached by 
a combination of methods, as follows : 
1. Direct. How commonly do we see clinical symp- 
toms and morbid anatomical changes that are incontro- 
vertibly due to the parasite? We restrict ourselves here to 
a narrow group of inf estments indeed, and think of such 
diseases as trichosomiasis in prairie dogs and spiroteria- 
sis in parrots. 
2. By comparison wdth analogous infestments of 
domestic animals and man — more thoroughly studied and 
therefore more accurately appraised, in general, as to 
pathogenicity ; a comparison from the standpoint of dis- 
ease production rather than natural habits of the para- 
site. Example, coccidiosis and hookworm disease in 
foxes and dogs. 
3. By inference through deduction. This is the most 
unsatisfactory consideration of all, and should be well 
checked up and discounted. Here we would evaluate the 
known propensities of the parasite first, such as its size, 
motility, anatomic position in the host and the general 
pathological traits of the genus and family to which it 
