115 
with the most numerous radials, and present with the fewest, such a form would 
have to be separated solely on the absence of the central spine, and even in the orig- 
inal description of impexicoma the central spine is only said to be “ sometimes want- 
ing.” It has been impossible for me to separate the forms. It should be said that 
the fruit and seed characters given above were taken [rom a specimen whose few 
radials and no centrals would undoubtedly refer it to impericoma. As yet we are 
ignorant of the flower of C. corniferus. Vor discussion of relationships see under C, 
scolymoides. 
++ ++ Central spines 1 to 4. 
48. Cactus scolymoides (Scheidw.) Kuntze, Rev. Gen. Pl. 261 (1891), 
Mamillaria scolymoides Scheidw. Allg. Gart. Zeit, ix. 44 (1841). 
Globose or ovate, 5 to 7.5 em, high, subsimple: tubercles conical, 10 to 
16 mm. long, the upper elongated, incurved and imbricate: radial spines: 
14 to 20, straight or often recurved, white or horny, 10 to 20 mm. long 
(the upper the longer); central spines 1 to 4, longer (18 to 62 mm,), more 
dusky, curved, the upper ones turned upwards and intermixed with the 
radials, the lower one stouter, longer, and curved downwards: flowers 
5 em. long: fruit unknown.—Type unknown. 
From the Pecos River, western Texas, westward into southern New 
Mexico, and southward into Chihuahua and San Luis Potosi. 
Specimens examined: TEXAS (Hays of 1858): New MrExtico (Bige- 
low of 1853): CuimuAUA (Wislizenus of 1846): also specimens cul- 
tivated in St. Louis in 1858. 
Specimens collected by Mrs. Anna B. Nickels across the Rio Grande from Laredo, 
Texas, and showing neither flower nor fruit, seem to intergrade between C. scoly- 
moides and C. scolymoides sulcatus. The habit is that of the former, the tubercles are 
those of the latter, while the spines are somewhat dillerent fromeither. The number 
of central spines in these specimens is very hard to determine, as on the adult tuber- 
cle they all assume a radial position. The usual adult arrangement is an apparent 
absence of central spines; 10 to 12 rigid, spreading and more or less recurved radials 
(increasing in length from the lowest), which are mostly white or the upper more or 
less dusky; and above, just behind the radial row, 2 or 3 stout recurved-ascending 
spines, which are white with tips more or less reddish-black, one of the spines usu- 
ally much stouter and longer than the others. This form may represent a distinct 
species, but it seems very unsafe to add species to the ©. scolymoides group without 
the fullest information. 
Prince Salm-Dyck refers C. scolymoides to ‘‘M. daimonoceras Lem, Cact. gen. nov., 
p. 5,” but no mention of such a name can be found in the work referred to. Labouret 
refers C. corniferus to the same name and reference. If ‘“‘ M, daimonoceras ” wasany- 
thing more than a garden or herbarium name used by Lemaire [ have been unable to 
find it, and Dr. Engelmann’s notes indicate that his search met with the same result. 
It is possible that the name was applied loosely to this assemblage of closely related 
forms that seem to cluster about C. corniferus, 
A most perplexing question of relationship is presented by the forms that have 
been called pectinatus, scolymoides, sulcatus (ealcaralis), Echinus, and the Mexican 
forms radians, impexicomus, and corniferus. It may be that they are all merely vari- 
eties of one strong polymorphic type, but our knowledge of corniferus is so incom- 
plete, and material of other forms is so scanty, that I can not venture to make such 
an assertion. However, it seems probable that radians, pectinatus, scolymoides, sul- 
catus, and Echinus all have green fruit, while in impericomus and corniferus it is red. 
It has also seemed proper to merge radians and pectinatus, also impericonus and. corni- 
ferus, and to refer sulcatus to scolymoides as a variety, These seven forms are thus 
reduced at least to four species. 
