92 SECTIONAL ADDRESSES. 
only two branches? Field evidence supplies the answer to this very 
natural query. Whilst the earliest graptolite with which we are 
acquainted was a pendent form, there very quickly followed other 
graptolites in which a horizontal direction of growth replaced the earlier 
pendent direction, though both occur side by side in rocks of the same 
age; the horizontal growing forms we term Clonograptus, the pendent 
Bryograptus. Now it is perfectly obvious from observation in the field 
that as regards simplification of branching the same plan of evolution 
was followed in both these groups, though there is always a tendency 
for development to lag behind and go slower in the pendent line, 
whereas development is so rapid in the horizontal line that many Clono- 
grapti persist alongside the 8-branched Dichograpti, though when 
Dichograpti persist alongside the 4-branched Tetragrapti they are most 
commonly those in which a certain amount of simplification has already 
taken place, since they are, as a rule, forms with only six or five stipes 
instead of eight ; owing, however, to the unequal rate of development in 
the two groups there is a characteristic association of pendent 4-branched 
graptolites with horizontal 2-branched forms of the type of D. ezxtensus, 
whilst horizontal 4-branched forms have become rare. 
The apparent anomaly is thus clear when followed out step by step. 
This greater rapidity of development in one group seems to indicate that 
these horizontal-growing forms were the more successful of the two; and 
it is, therefore, perhaps only to be expected that almost all the later 
graptolites are developed from ancestors within that group, the excep- 
tions being those whose ancestry is at present obscure, but there is no 
indication that these arose from any member of the pendent group; I 
have so far been utterly unable to find any graptolites in later beds 
which seem to be connected with these. If I am right in supposing 
that the change in direction of growth of the rhabdosoma was connected 
with the protection of the hollow thread-like nema (virgula auctorum), 
which is the attachment organ so vitally necessary to the colonial organ- 
ism, the forms belonging to the pendent group may be regarded as 
unsuccessful because they fail utterly to secure this necessary protec- 
tion, and, therefore, the members of this group come entirely to an end 
at the top of the Llanvirnian. Within the other group protection is 
better achieved, since in many cases at any rate the horizontal-growing 
stipes appear to have been plastered on to foreign bodies or suspended 
therefrom by short threads, and the nema would, therefore, in most 
cases have been short. Within this horizontal group the goal in simpli- 
fication would seem to have been reached early in the one-stiped A zygo- 
graptus of the Middle Arenig; this type is repeated more than once at 
slightly higher horizons, but appears to be in no case a successful form; 
individuals are very commonly broken in the region of the sicula, which 
is in itself suggestive, and they, like the pendent Didymograpti, appear 
to be ‘dead ends.’ The successful one-stiped form is attained much 
later by very devious routes through the scandent or climbing grapto- 
lites, and in all of them the attachment organ is very perfectly pro- 
tected, partly by being buried within the rhabdosoma for the whole of 
its initial region, and partly by the development of a special encasing 
tube or sheath, 
