150 Mr. H. M. Bernard on 



that these particular scars are the segmental repetitions of 

 the stigmatic apertures, whether they are now anything more 

 than scars or not. Hansen, however, points out that similar 

 scars may be found on the anterior abdominal segments 

 nearer the middle line concurrently with those that repeat 

 the stigmata. This interesting fact, which I have since 

 fully confirmed for two segments (II and III) in my original 

 specimen, in no way affects the above conclusions. There 

 must at one time therefore have been apertures here also 

 which have now closed. It is hardly likely that they were 

 tracheal invaginations, although there is no impossibility in 

 there being two pairs to one segment — that is, if tracheal 

 invaginations can be deduced from primitive setiparous glands. 

 It is more probable that they were the openings of spinning- 

 or cement-glands. Such glands exist in these very segments 

 in many Chernetida? ; and in the Araneidae, in addition to 

 the large spinning-mammillae, smaller ones may occur in 

 the same segments nearer the median line. We are, how- 

 ever, here chiefly concerned with the fact that a row of scars 

 segmentally repeats the functional stigmata along the whole 

 length of the abdomen. 



The bearing of this on the origin of the Arachnida I have 

 already discussed in ' Nature' *. It points to the deduction 

 of the Arachnida from an ancestral form with a pair of limbs 

 and a pair of stigmata on every segment. 



Further evidence of this has been slowly accumulating. 

 Reserving for the present that which is specially connected 

 with the Galeodida?, one important item deserves to be sepa- 

 rately discussed. If any collection of Thelyphonidse be looked 

 through carefully f it is impossible to avoid the conclusion 

 that these Arachnids once possessed limbs with stigmata 

 along at least seven abdominal segments. The specimens 

 require to be dried and then held so that the abdomen reflects 

 the light from the window. Very few indeed showed no 

 traces at all. The large majority show on segments V, VI, 

 VII, VIII, just laterally to the muscle-impressions, definite 

 scar- like markings, or even sharply circumscribed areas, such 

 as I have shown in figs. 1 and 2. 



The posterior margin of this area is, as a rule, seen most 

 distinctly. In some, however, the strongest mark is the inner 

 posterior corner of the area; in others, again, the inner 

 longitudinal side, which, however, then generally slopes 

 outward posteriorly (e. g. many individuals of Theli/phonus 



* " The Stigmata of the Arachnida as a Clue to their Ancestry," ' Nature,' 

 Nov. 16, 1893.' 



t My hest thanks are due to my friend Mr. R. I. Pocoek for permitting 

 me to examine many scores of specimens (both alcohol and dry) under his 

 charge in the South Kensington Museum. I was enabled to examine 

 specimens of five genera — Thehjphonns, Thehjphonellus, Uroproctus, Masti- 

 gopi-octus, and Typopeltis. 



