safford: myrmecophilous acacias 357 



number of well defined natural groups. When it came to his 

 notice that Dr. Schenck was also engaged in studying the myrme- 

 cophilous Acacias, the writer postponed the publication of his 

 paper in deference to Dr. Schenck, from whom he received a 

 most courteous acknowledgment. ^ 



Dr. Schenck based his classification principally upon the 

 venation of the leaflets and upon the general form of the inflores- 

 cence. In his material fruits of several species were lacking. 

 On the receipt of photographs of further specimens of his newly 

 established species, many of which included seed pods, Dr. Schenck 

 realized the importance of using the latter as a basis for natural 

 classification.^ As it has become necessary to defer the publica- 

 tion of the wTiter's monograph of the myrmecophilous Acacias, it 

 is thought advisable to offer the present preliminary paper, in 

 which a classification of the group is presented together with 

 descriptions of nine new species. 



KEY TO THE GROUPS 



Involucre situated near the base or at least below 



the middle of the peduncle; interfloral 



pedicelled bracteoles peltate; spines 



glabrous 



Pericarp indehiscent, inflated, terminating in 



an acute spine-like beak I. Ceratophysae. 



Pericarp dehisc6nt 



Fruit a pod, opening b}-- a dorsal and a 

 ventral suture 

 Flowers in globose heads; pods some- 

 times very long and slender .. . II. Globuliferae. 

 Flowers in cjdindrical spikes; pods 



short and relatively broad III. Leguminiferae. 



Fruit a follicle, opening by a single suture. IV. Folliculares. 

 Involucre situated at or above the middle of the 

 peduncle, the latter pubescent or hirtel- 

 lous; interfloral pedicelled bracteoles not 

 peltate ; spines puberulent when young ... V. Hebacanthae. 



-See H. Schenck, "Die myrmekophilen Acacia-Arten," Bot. Jahrb. Engler 50: 

 449-487. 1914. 



' "Die BeschaffenheitderFriichtedlirftevielleicht zur endgiiltigen Unterscheid- 

 ung der Artengruppen der Ameisenakazien mindestens ebenso wichtige, vielleicht 

 noch bessere Anhaltspunkte geben wie die Form der Inflorescenzen." Bot. 

 Jahrb. Engler 50: 480. 1914. 



