412 ANNUAL, BEPORT SMITHSONIAN INSTITUTION, 19 3 



scales increase in number and decrease in size from the root attach- 

 ment to the top of the plant, changing from a brown color under- 

 ground to a light purple as the head, or disk, is reached. The number 

 of stalks branching from a single contact on the host root varied from 

 one or two to as many as 25, but the usual number was from four to 

 eight. 



Rarely did the stalks ascend perpendicularly from the contact 

 point with host root, usually curving horizontally and then upward 

 at various angles. The heads Avere sometimes close together at the 

 surface, but usually as much as 3 or 4 feet apart. In some in- 

 stances we found the stalk only part way to the surface and then, 

 instead of a head, it had a sharp point for penetrating the sand. In 

 one instance this sharp point had grown entirely through another 

 Ammobroma stalk at right angles to it. 



The young, tender stalks were slightly enlarged soon after leav- 

 ing the host root, and some of them showed an enlargement just 

 below the surface of the sand. Many of the stalks had thrown out 

 a few roots at the base, or point of contact with the host root, thus 

 indicating an attempt to depart from pure parasitism. In a few 

 cases, as shown by the photographs, there were many Ammobroma 

 roots. When the stalks were in their prime they Avere very succu- 

 lent, plump, and very tender and brittle, but as the moisture disap- 

 peared and the heads matured, the stalks Avilted, decreased in diam- 

 eter, and became flabby. Apparently the head forms just before it 

 reaches the surface of the sand and then pushes its way up to the 

 surface in a manner similar to the growth of a puffball or toadstool. 



When moisture is available under the conditions prevailing in the 

 sand dunes of Imperial County, Calif., this parasite is able to store 

 sufficient moisture and nutriment in its stalk to continue flowering 

 well into the rainless summer months. We collected our largest and 

 best heads on April 26, 1928, at a time Avhen no moisture could be 

 detected in the sand at the point of contact with the host and the 

 roots of the parasite appeared to be dead. The Ammobroma stalks 

 Avere wilted, indicating that the heads had continued to grow and 

 mature seed from tlie storage in the stalks. 



The disproportion betAveen the small nonsucculent Coldenia host 

 and the heavy, fleshy and succulent parasite is very noticeable, the 

 parasite often being many times the weight of the host. This is 

 possible because the Ammobroma has its OAvn root system, enabling 

 it rapidly to absorb Avater Avhen available in the sand of the dunes. 



The host plants of Ammobroma were not identified by Col. A. B. 

 Gray, who discovered it near Adair Bay, Sonora, Mexico, on May 

 17, 1854, nor by Carl Schuchard, who in May, 1858, between Pilot 

 Knob and Cook's Well, collected the specimens Avhich were studied 

 so carefully by Count Solms-Laubach. 



