i:; 



The minute, subulate, early deciduous leaves of this genus furnish the 

 nearest approach to true leaves found among our native species. 



In all the specimens examined, true epidermal and hypodermal regions 

 occur in sharply defined outline. The cuticular layer is generally thick- 

 ened and is clearly distinguished from the true epidermal walls. It be- 

 comes thinner as it nears the stoma, and is easily traced into the air 

 chamber (?) It completely lines this respiratory cavity, and, as Yon Mohl 

 shows, it even sends out open tubes into the adjoining inter-cellular 

 spaces. The range in thickness passes from the very thin, almost imper- 

 ceptible form seen in Mamillaria macromem to the astonishing thickness of 

 that seen in Mamillaria (Anhalonium) prismatica, where the cuticle is fully 

 ten times the thickness of the true epidermal layer underneath. The 

 stoma in this species communicates with the outer air by a chimney-like 

 canal extending upward and outward through the cuticle. This canal or 

 chimney is beset at three different elevations by sets of four Hap like pro- 

 jections which extend out from the wall in such manner as to almost en- 

 tirely close the orifice. I have failed to find anywhere any mention of 

 these projecting appendages, but conclude that their function is undoubt- 

 edly that of accessory guard cells of the doma. They readily expand on 

 application of moisture, which fact in itself is sufficient evidence of their 

 purpose. The outer wall of the true epidermis in this species barely 

 reaches an average development ; while the hypodermal region consists of 

 but one layer of moderately thick-walled narrow cells. The only remain- 

 ing feature of the cuticle worthy of note as a diagnostic character is the 

 undulation of surface, which is displayed in certain species. Prominent 

 elevations occur in Cenus Greggii, Cereus horizonthalonius, var. centrospinus, 

 and in Echinocactus polycepJialus.. 



Definitive characters in the true epidermis are not abundant ; but, when 

 they do occur, they are distinct and unmistakable. Three species of 

 Opuntia show tangential (?) partitions in the epidermis, breaking it up 

 into two or more rows. Opuntia phscacantha has its epidermis thus 

 thrown into eight rows of exceedingly thin -walled cells. A new species of 

 cereus (as yet unpublished) has as high as nine rows of this thin-walled 

 epidermis. Species having two layers are Echin. pob/ccphalus, three un- 

 named species of Cereus from San Louis I'otosi, Mex., and a new species of 

 Cereus from Casa Grande, Ariz. The most curious epidermis in the entire 

 collection is that of Echin longihamatus. Since there is nothing like it in 

 the entire number observed, it is well deserving of more than passing at- 



