316 



MEMOIRS OF THE NATIONAL ^OADEMY OF SCIENCES. 



extend outward and downward as two simple, slender tubes or ribbon-like filaments. In the second 

 form tliey divide in the same manner, but the two branches are spirals. (See Fig. xvii.) 



The third type, which is apparently confined to the anterior segments of the thorax, is figured 

 in his PI. Ill, Fig. 2. Here there was a series of at least 17 thin leaves arranged from above 



Fig. X VIII._A, a leg of Branchinecta lindahli (from 10th pair) : <7', 9''. giUs ; P, P', the 

 two parapoiliiuu-liko lobes, comparable with those of the annelid Lyaret»(B), rfc, 

 dorsal cirrus, acting as a gill. 



Fig. XIX.— Leg oi 

 glomerans. with : 

 tral cirrus (c')div 

 Ehlers. 



parapodium of Eunice con- 

 . branched gill and the ven- 

 idfd into two joints. — After 



downwards, not spiral but flat, somewhat as in Limulus, and radiating vertically from a jointed 

 i'l)ipodite or biancliial arm. Thus as in forms like Calymene, Asaphus, Brouteus, etc., all the 

 limbs behind the mouth-region were cylindrical, ambulatory, and branchiferous, but those of the 

 pygidium were not broad and flat and adapted for swimming, as in Limulus. Thus the Trilobites 

 were less differentiated and more closely allied to the Annelid tyjje than the Eurypterida and Lim- 

 ulus. It thus seems probable that the three great groups of Arthropoda had each an independent 

 origin from the higher worms, or ft-om extinct forms like them. 

 The insects, with the Myriopods, were independently evolved 

 fiom some Cha^topod worms of unknown afiflnities by way of 

 Peripatus; the Podostomata* by an independent path also 

 originated from a true branchiate Annelid, while the Crustacea 

 with litile doubt, as generally conceded, arose from some 

 Annelid like form; perhaj)s one with dorsal cirri acting as 

 branchiae, like those of Lyareie braziHensi.s, figured by Ehlers 

 (Mem. M. C. Z., 1887), these being remarkably like those of 

 Branchipus and other Phyllopods.t (Fig. xviii, B.) 



It is also to be borne in mind that in the trilobites, the 

 branchise were dorsal though lying under the dorsal wall of 

 the body, and attached to an epipodal division of the limbs, 

 and so they are in Limulus, as shown by our Fig. xx. This dorsal position of the gill-plates should 

 be taken into account when comparing them with the book-leaves of Arachnids. 



Fig. XX. — Set-tion through the abdomen of Limu- 

 lus; c, carapace: ht., heart; int., intestine; ng., 

 ganglion ; en, axial jointed endopodite ; ex epipo- 

 dal portion bearing the gilla. 



* The class Podostomata includes l>oth the Merostomata, of wliicli Limulus is the living rej)resentatiTe, and the 

 Trilobites. 



+ The problem of the derivation of the jointed limbs of Arthropods from the soft non-jointed Annelid para- 

 jiodium is paralleled by that of the origin of the limljs of the higher vertebrates from the fins of fishes. The limbs 

 of Arthropods are divided into a fevr or many leverage systems, represented by segments or joints. There are all 

 grades of Arthropod limbs, from the uujointed parapodia-like limbs of Phyllopods, which are most Annelid-like, to 

 the imperfectly jointed swiijjming legs of the Nauplius of various Crustacea, and to the abdominal legs of Slalacostraca, 

 u]) to the amliuUitory legs of Decapods and of terrestrial tracheate Arthropods. These v.arieties of shape can be ex- 

 ]il:iiiH-il liy .iihipliitiou to changed conditions resulting from a change from an aquatic to a terrestrial life, or from 

 Hiiii]ily .swiiiiiiiiiii; to walking on the bottom or to burrowing in sand and mud. On the other hand the appendages of 

 .Syllid Avmiris arc often jointed, and in Eunice (Fig. xix) and Polynoe, the lower division of the parapodium con- 

 sists of two joints. 



