406 Dar bishire. — Observations on Mamillaria elongata . 
we primarily look at these large cells as an arrangement for the reduction 
of transpiration. Before however any opinion could be expressed on this 
question, it would be necessary to determine more carefully the histology 
of these epidermal cells, and the exact nature, chemical composition and 
degrees of concentration of their contents. It might then be possible 
to explain their relation to transpiration more fully. As it is now, 
I consider that they represent paraheliode structures. The rate of trans- 
piration is reduced by the large epidermal cells keeping down the 
circulation of air around the stomata, a fact which was proved experi- 
mentally by Hagen some thirty years ago (14, p. 24). The stomata are 
placed in between these peculiar cells and do not lie exposed to the air 
directly. 
Each leaf is crowned by a set of hairs. Each hair is derived from 
a single epidermal cell. But despite its unicellular nature, it very much 
resembles in structure one of the multicellular spines of Mamillaria 
elongata. It has a pointed upper end, but its lower end is much swollen 
and contains a glistening mass of air. The swollen part is also coloured 
brick-red, and will absorb rays from the sun. These hairs are very soft 
and almost papery in texture, but I have no doubt that here, as in 
Mamillaria elongata , they form a useful sunscreen or paraheliode for the 
protection of the underlying leaf-structures, the endings of the vascular 
tissue being found most abundantly here. 
Mesembryanthemum stellatum forms, I think, a very strong parallel 
case in support of my views on the function of the spines, as, though being 
a member of a natural order quite different to the Cactaceae, it has never- 
theless developed the same plant-form as these. 
A very great number of Cactaceae belong to the same plant-form as 
Mamillaria elongata. Schumann has pictured many of these in his Icono- 
graphia Cactacearum. His representation of Echinocereus subinermis , 
Salm-Dyck (27, Vol. I, PI. Ill), is very interesting on account of the fact 
that the young flower-shoots are well protected by spiny paraheliodes, but 
in the older portions these almost disappear (26, p. 250). The tubercles, 
with the apical set of spines, gradually pass into the outer leaves of the 
flower. These leaves are foliaceous in form, and they also at first bear an 
apical set of spines. This confirms my view on the homology of the spines 
in Mamillaria elongata. 
Paraheliode structures of the nature and kind described for Mamillaria 
elongata are not, however, met with in all Cactaceae. It would be very 
interesting and most instructive to examine a large number of species 
of this natural order, and determine the relation existing between the 
external conditions on the one hand and the development of the parahe- 
liodes on the other. It would be of great interest to determine whether, 
when the paraheliode spines are smaller and evidently less efficient as 
