CONTRIBUTIONS TO WESTERN BOTANY NO. 15 47 



black and! about 4-6 mm. long and 3-4' mm, wide. The form Sonorae, as it 

 grows in the Cape region of lower California is mostly staminate, rarely 

 does one see a pistillate tree, or even a monoecious one. The leaves of all 

 forms of filifera hang on for many years as a rule, but there are occa- 

 sional exceptions where they fall off quickly and leave the trunks bare, at 

 least in forms of Sonorae. A character that is common to all the palms 

 is the production of subcutaneous rootlets which swell out the bark at the 

 base of the tree and finally make it split as in Phoenix. A^'hcn the leaves 

 of filifera have fallen naturally they leave the trunk with very slight rings 

 which are not sharp on the edges, as in the cocoanuts- Watson in his 

 characterization of Erythea makes woolly spathes a generic character, as 

 well as perfect flowers. The spathes of E. edulis are woolly, those of K. 

 armata are perfectly smooth. The flowers of E. edulis are evidently 

 polygamous or monoecious as are those of E. armata at least at times, bui 

 there are dioecious trees. The flowers of both species of Erythea ar^* 

 greenish white and 2-4 mm. long. The fruit of E. edulis is blu^k and 

 edible and about an inch wide. That of armata is half as large. The 

 leaves of edulis fall off in 2-3 years, leaving sharp rings around the 

 trunk, and exposing very conspicuous sheaths- between the leaves which 

 are about a foot wide. In E. armata these sheaths are less than half a^ 

 wide but still conspicuous. The leaves af armata hang on for years as in 

 W. filifera. The threads of filifera are not so conspicuous in Er)'thea a> 

 in most Washingtonias. The petrioles of E. edulis often have minute 

 prickers, which seem deciduous. Those of E, armata are like those of 

 Washingtonia filifera. There is not a character proposed by anyone for 

 Erythea to separate it from Washingtonia that has any value at all, but 

 there seems to be a character of value in the flowers (see beelow). 



I can see no sound foundation for Cooke'^ genus Glaucothea, "Tiie 

 flowers are the same as in Erythea edulis. The fruits are intermediate 

 between Erythea and Washingtonia, and the leaves and habit are oi 



Washingtonia- 



It has been my privilege to see the cultivated forms of Washingtonnv 

 in southern California. I have also visited all the large areas where var. 

 rubusta grows wild, and have also studied it in cultivation all over the 

 region. I have also studied the genus all along both coasts of Lower 

 California and on the mainland as far south as Guadalajara, throu;:"h 

 Sonora, Sinaloa, and Nayarit, and at the Cape. In Mexico the plants 

 can hardly be said to be cultivated, except in the Mexican stupid way^ 

 having grown up along watercourses haphazard, the original seeds prob- 

 ably having been scattered by man after obtaining them from the canon? 

 of the adjoining mountains, and for the purpose of having the leaves 

 nearer at hand for thatch, rope, and baskets. The Mexicans distinguish 

 forms by the color of the wood, calling them *Talma colorada and Palma 

 blanca, and Palma negra. W, filifera goes by the names palma blanca 

 and palma colorada. Erythea armata and Brandegei are called palma 

 negra. 



The only form that grows wild in California is Washingtonia filifera 

 var. robusta (Wendh Gart. Zeit 2 198 1883 as species). Parish Bot 

 Ga2. 44 420 1907. It is a well marked form, and the best developed oT 



