SEGMENTATION AND PHYLOGENY OF ARTHROPODA, 483 
number of their trunk-segments (see Table, p. 472), but that 
there is a missing head-segment in all Arachnida except the 
Pycnogonida. The only embryological evidence for such a seg- 
ment is afforded by the evanescent somite with vestigial limbs 
described by Lendl (22) in the embryo of the spider Epeira, 
between the somite of the chelicerze and that of the pedipalps. 
Unfortunately, Lendl gives no figures, but his description is 
precise, and the late appearance of the head-segments in the 
embryonic development of Arachnids generally makes it easy 
to believe that a segment has disappeared from that region. 
The suggested homology between the appendages of this lost 
seoment and the palps of a Pycnogon brings the four pairs of 
walking-legs among the Pycnogonida into correspondence 
with those of spiders. However, Hodgson’s (12) recent dis- 
covery in the Antarctic seas of a Nymphonid, and his re- 
discovery of a Collosendeid (18) with five pairs of legs may 
be thought to render this comparison of little value; for if 
Pentanymphon and Decalopoda be primitive forms, we must 
consider the Pycnogonida as probably descended from an- 
cestors with many pairs of similar legs, most of which have 
disappeared successively from behind forwards. But the fifth 
pair of legs in these genera may possibly represent a com- 
paratively new development, and, in either case, if the view 
now advocated be accepted, we find that the head of a 
Solifugid, of a Pycnogon, of a Trilobite, and of a typical 
Crustacean correspond exactly. Our conclusion is, there- 
fore, that the Azachnida must be traced back to proto- 
Trilobitan ancestors which possessed a head with five and a 
trunk with fifteen limb-bearing segments. 
As to the relationships between the orders of Arachnida, 
my views already published agree in all essential points with 
those of Lankester and Pocock (20), except that I would 
assign to the Solifugida and the Palpigradi a more primitive 
position than they do in the scheme of arrangement. The 
free thoracic segments of those animals forbid us to derive 
them from a scorpionoid form in which the segments of the 
cephalothorax had already become aggregated. In their 
