Tree (Ginkgo biloha , L.). 135 
and assumes the form of a ring consisting of seven or eight 
collateral bundles of xylem and phloem. 
In a fairly old root the annual rings are not so well marked 
as in the stem, and the walls of the tracheids are thinner ; the 
medullary rays which contain a few crystal-sacs vary in 
height from one to seven cells. The xylem parenchyma in¬ 
cludes some crystal-containing cells, and thick bast fibres are 
abundant in the phloem. 
FOSSIL GINKGOACEAE. 
To attempt a detailed description of the fossil records 
which have a direct or indirect bearing on the geological 
history of the Ginkgoaceae would necessitate an unreasonable 
extension of the present paper. Our aim must be confined 
to giving, as briefly as possible, a general summary of the 
more trustworthy evidence which may enable us to form an 
opinion as to the antiquity of the Ginkgoaceae, and as to the 
geographical distribution of fossil species in former periods of 
the earth’s history. In 1881 Oswald Heer 1 published an 
interesting article, in which he summarized the data furnished 
by Palaeozoic, Mesozoic, and Tertiary fossils towards the past 
history of the existing type. The list of plants enumerated 
includes certain forms which cannot be satisfactorily shown 
to bear a close relationship to Ginkgo , and, on the other hand, 
subsequent research enables us to make additions to Heer’s 
list of extinct species. 
Palaeozoic leaves. Among the genera recorded from Car¬ 
boniferous and Permian rocks, which have been regarded as 
more or less closely related to the existing Maidenhair tree, 
we have Ginkgo itself, Ginkgophyllum , Baiera , Saportaea , 
Trichopitys , Dicranophyllum , Rhipidopsis , Whittlesey a, Psygmo - 
phyllum , Cordaites , Gomphostrobus , Trichophyllum , and others. 
Ginkgo (Salisbnria). Under the name Salisburia primi - 
genia , Saporta and Marion 2 have figured some fossils de¬ 
scribed by Grand’Eury from the middle Permian of the 
1 Heer (’81). 2 Saporta and Marion (’85), p. 145 , Fig. 74 . 
