1960] 



SPATHIPHYLLUM 



7 



DEVELOPMENTAL TRENDS 



The species of sect. Massowia appear to represent a phylogenetic offshoot of 

 great age. The union of their perianth segments into a 4—6-sided cup probably 

 was an evolutionary advance of ancient occurrence, since each of the three 

 species with this characteristic occurs in an area widely separated from the others. 

 8. cannaefolium, the most widespread of all the species of Spathiphyllum, is 

 restricted to northern South America east of the Andes, and is the only species 

 on the island of Trinidad. S. laeve is restricted to Cocos Island in the eastern 

 Pacific. S. commutation occurs in the same latitude as 8. cannaefolium, but on the 

 opposite side of the Pacific Ocean, where it is distributed on several islands in 

 Indonesia and the Philippines. These species are remarkably similar, and 

 surely must be monophyletic in origin. 



8. humboldtii, the sole species of sect. Dysspathiphyllum, appears to be a 

 variable species of wide range in South America. It is an enigmatic group of 

 obscure alliance. Its flowers are frequently reduced to a bimerous condition, while 

 the locules of the ovary are mostly multiovulate. There is a remote resemblance 

 between this species and 8. cuspidatum and 8. gardneri (here placed in sect. 

 Amomophyllum) . Possibly these three species arose from a common ancestral 

 stock, or perhaps hybridization has been a prime factor in their derivation. 



Speeies of the sections Spathiphyllum and Amomophyllum occur in Colombia 

 in close contact. This investigation has revealed little to suggest that one of these 

 groups developed from the other. It is more probable that they arose from a 

 common ancestral stock that no longer persists. That stock may have possessed 

 characteristics similar to those of 8. quindiuense, and perhaps also gave rise 

 to the species of sect. Massowia. 



There is- clearly a progressive reduction in the number of ovules per locule in 

 species of sections Spathiphyllum and Amomophyllum, which trend is correlated 

 with a reduction in size of the plant in the latter section. This reduction has 

 guided me in arrangement of the species within each section. 



The diagram of relationships (Fig. 3) suggests the possible origin and inter- 

 relationships of the species. 



COLLECTIONS AND TYPIFICATION 



The most important single assemblage of aroids, deposited at Vienna where 

 Schott had worked, was totally destroyed by fire in 1945. Nearly as serious was the 

 loss ( through war action) of many aroid collections at Berlin, where Engler had 

 worked. Most species of Spathiphyllum whose types were destroyed may be 

 recognized from their descriptions and illustrations. 7 A few, however, have been 

 difficult to distinguish because of the lack of definitive material, especially 8. 

 lanceae folium, the type species of the g'enus. It has seemed appropriate to 

 designate neotypes from some of these species. 8 For others the original descrip- 

 tions are considered as types, 9 and a representative collection is cited to exemplify 

 my interpretation of each. 



7 Excellent illustrations accompanied the original descriptions of many species of this 

 genus; at Vienna original plates by Schott are preserved that illustrate many species; for two 

 species there are authentic photographs of the type specimens made before those specimens 

 were destroyed. 



s International Code of Botanical Nomenclature (1956), Art. 7. 

 & Op. cit., Art. 10. 



