8 Memoirs Bcniicc P. Bishoj^ Miisciiiii. 



less isolated than now. During that epoch the fruits of the primitive i'riichartlias, 

 and also of other tropical plants, ma}- ha\e been deposited here and there on the 

 islands of the gri)Ui). by birds, ocean currents or even winds, thus creating a special 

 vegetation: Init when the great seismic and erupti\-e period arri\-ed, the vegetation 

 which covered the plains may have been carried, in part at least, on the mountains 

 which were rising, and to a certain extent may have been preserved upon them. 

 With this hypothesis a part of the vegetation now existing on the Hawaiian moun- 

 tain.^ could be considered as a surviving element of that which covered their |>lains 

 before the advent of the great cataclysms which compleleh- changed the orogra])h- 

 ical structure of the region. 



FERTILIZATIOX. 



The flowers of the Pritchardias open in an unusual manner. Their corollas 

 apparently perform one function only, that of sheltering the sexual organs. They 

 serve no vexillary purpose; that is to say, of attracting insects or birds. The flow- 

 ers are small, and only by their ensemble may attract those creatures who know 

 their meaning. The three segments of the corolla become detached from the 

 tubular portion projecting from the calyx, and drop at anthesis, that is when the 

 anthers are ready to scatter their pollen grains." This mode of flowering is per- 

 fectly analogous to that of certain Ampclideae and .Iraliaeeae. The petals having 

 fallen, the anthers, borne on short filaments, radiate from the mouth of the tube 

 of the corolla and around the style, which projects more or less from the stamens. 

 This contrivance seems to be very favorable to the impollination of the ovaries 

 In- bees or other flower-haunting insects, as in / If is and Hedcra. Perhaps also 

 as in the flowers of these plants some special odor, imperceptible to our senses, 

 serves to attract insects from a distance. 



I have no positive observations in regard to anv saccharine secretion in the 

 flowers of the Pritchardias, as it is not easy to verify this occurrence from dry 

 herbarium specimens. To tell the truth, though the tube of the corolla of . the 

 Pritchardias is of such a crjuformation that it might be taken for a recipient fitted 

 for collecting nectar, and e\'en its side walls, which are somewhat thick, would sug- 

 gest a nectarifluous structure, yet in all the llowers examined I have not met with 

 anv trace of a saccharine secretion. 



The pollen grains of Pr. affiiiis, which 1 have examined v.ith the microscope, 

 are. in the dry state, ellipsoidal in shape with a deep longitudinal fold, but when 

 swelled bv immersion in water, assume an ovoid-ellipsoidal or even subglobose 

 shape; their surface is perfectly smooth, without asperities or prominences. Never- 

 theless the grains seem to have a tendency to agglutination, and a]ipcar to belong to 

 the type of pollen grains best adapted to be collected by insects adhering to the body 

 of any animal which might be in the habit of frequenting flowers, rather than to the 

 type of pollen grains easily transported by winds. This appearance, which I have 



sMalesia, vol. II, t. XXXVIII, f. 1-3. 



