544 ANNUAL EEPOET SMITHSONIAN INSTITUTION, 1908. 



one another, as in the structure of the seeds, yet in some cases there 

 seem to be intermediate forms, and the flowers themselves are remark- 

 ably similar. Notv/ithstanding this, it would be of gi'eat assistance to 

 the student of Cactacese to recognize many of the so-called subgenera 

 as genera^ not only for convenience of study, but also to show the 

 systematic connection between the members of each group. In assign- 

 ing the various species to a particular genus it will often be difficult 

 to decide to which of two it belongs, but as Schumann says, this 

 difficulty can not be avoided unless all the Cactacese be combined 

 into a single genus Cactus.^ 



In the following synopsis of Mexican genera, while following Schu- 

 mann's arrangement I shall include various genera of Opuntioidea3 

 and Cereoidese established by Britton and Eose, including Pere- 

 skiopsis and Camegiea^ and the classification of Cereus and its allies 

 published by Mr. Alwin Berger in the Sixteenth Eeport of the Mis- 

 souri Botanical Garden, 1905, and by Britton and Rose in the Contri- 

 butions from the United States National Herbarium, volume 12, 

 pages 413 to 437, 1009. I have also consulted Labouret's Monographic 

 des Cactees (Paris, 1858) ; Engelmann's Cactacese of the Mexican 

 Boundary, with its beautiful plates by Paulus Roetter (Washington, 

 1859) ; Lemaire's delightful little handbook, Les Cactees (Paris, 1868) ; 

 Forster's Handbuch der Cacteenkunde, second edition (Leipsic, 1886) ; 

 Coulter's Botany of Western Texas (Washington, 1891-1894) and his 

 revision of the North American Cactacese in volume 3 of the Contri- 

 butions from the United States National Herbarium (Washington, 

 1894r-1896) ; Schumann's Gesammtbeschreibung der Kakteen (Neu- 

 damm, 1899) ; and various numbers of the Monatsschift fiir Kakteen- 

 kunde. The plates are reproductions of photographs by Mr. Guy N. 

 Collins, Mr. C. B. Doyle, and Mr. E. L. Crandall, of the United States 

 Department of Agriculture, Mr. T. W. Smillie, of the United States 

 National Museum, and myself. The text-figures were drawn by 

 Mr. Theodor Bolton, principally from photographs, and from the 

 plates of Paulus Roetter already referred to. 



Synopsis of Mexican Cactace^. 



i. subfamily pereskioideie. 



The best known representative of this subfamily, which consists 

 of the single genus Pereslda, is the Barbados gooseberry Pereskia 

 aculeata, a scrambling shrub with long, slender branches armed with 

 recurved prickles, and glossy green leaves^ like those of a lemon. 

 The flowers grow in clusters and are peduncled, unlike those of all 



« Scliumanu, Karl. Keys of the Monograpli of Cactaceae, p. G. 1903. 



