XXXVI HABITS AND STATIONS. 



The presence of special color-patterns on single trees or 

 small groups, in many instances, is an evidence of the slow- 

 ness of migration, as well as of the facility attending the 

 origin of varieties. 



Food.- -The frequent presence of living snails on trees long 

 dead indicates that their food is cryptoganic. Moreover, the 

 leaves show no traces of grazing. The feces of Partulina 

 confusa examined by Mr. Henshaw were composed chiefly of 

 remains of fungi or algae (p. 103). The contents of stomachs 

 of Achatinella mustelina and several other species, examined 

 by Dr. A. Brown and myself, were recognized as fungi, often 

 with shreds or fibres probably of bark, which remained un- 

 digested in the intestine. 



AchatinellidcF breed at all times of year. In any catch one 

 finds part of the individuals carrying embryos. The length 

 of life of the individual is not known. 



The native Hawaiians claim that the tree shells have a 

 song, which they have fancifully supplied with words. Dr. 

 Newcomb (P. Z. S. 1853, p. 129) and Dr. N. B. Emerson 

 (Sacred Songs of the Hula, p. 121), and others have given 

 versions of this song. Mr. Perkins believes it to be the chirp- 

 ing of crickets. 



Brief notes on the topography and conditions of collecting 

 may be useful to conchologists who have not visited these 

 islands. Those interested in the subject should consult the 

 Introduction of Fauna Hawaiiensis by Mr. B. C. L. Perkins. 



The Achatinellge are not shells of the valleys but of the 

 ridges and upper ravines. The bottoms of the larger valleys 

 often lie below the zone of requisite humidity. This is doubt- 

 less much more generally true now than before the valleys 

 were so extensively deforested. Certainly at the present time 

 the ridges and their slopes are the chief natural highways of 

 the tree-snails, and the valleys are barriers, in varying de- 

 gree, to the spread of species along the ranges. 



Since it is the valleys and not the ridges which figure in 

 topographic nomenclature, it follows that the usual locality 

 records by valleys are often less definite than we would de- 

 sire. For instance the locality "Nuuanu valley" might mean 



