ZOOLOGY AND BOTANY, MICROSCOPY, ETC. ^589 



ectodermal parts of tlie segmental organs to their mesodermal rudiment. 

 As, however, there are two canals in B. Kuppferi, we may homologize 

 their duct with the cephalic segmental organs (but not with head- 

 kidneys). 



The folds of the inner peritoneal layer have not, as Bateson thought, 

 anything to do with the so-called proboscis gland, but rather have the 

 same kind of relation to the vascular system as have the pericardial 

 glands of Annelids ; their function is that of excretory organs. The 

 proboscis gland of Bateson (heart of Spengel), has a proper muscular 

 wall, while its epithelium is like that of the endothelium of some parts 

 of the peritoneum of the proboscis. If S^jengel is correct in regarding 

 this organ as derived from the pulsating vesicle of Tornaria, it may be 

 compared with the similar vesicle of Molluscan larvae. These last (e. g. 

 in Limax) do not communicate with the vascular system, but with the 

 cavity inclosed between the two mesodermal layers, which Salensky, in 

 Vermetus, homologized with the coelomic cavity. The organ which 

 Bateson regarded as the notochord cannot be compared witli the true 

 chord ; it probably represents the preoral part of the enteron, and is 

 well developed in correlation with the great development of the preoral 

 lobe. The lacuna in the proboscis appears to have no proper walls, and 

 must not be homologized with the heart. 



The vascular system of this species appears to be very simple, the 

 dorsal and ventral trunks alone being developed. The layer of nerve- 

 fibres is under the whole of the integument ; its dorsal and ventral 

 thickenings have a simpler relation to the body epithelium than is the 

 case in B. minutus. The dorsal central nervous system has no central 

 cavity, no neuropores, and no dorsal cords. 



The skeleton, as Spengel rightly supposed, is merely a local thicken- 

 ing of the membrana propria, and has none of the morphological 

 elements described by Marion in B. Talahoti. The branchial part of 

 the enteric canal forms a few loops, and the gills have the same structure 

 as in B. KoioalevsJcii. The branchial region divides into two parts — an 

 upper with the epibranchial ridge, and a lower which has the form of a 

 small groove, the base of which is beset with papillaB. This groove is 

 similar to the diverticulum of the segmented portion, and the two may 

 be regarded as the homologue of the endostyle, the hypobranchial 

 groove, and the thyroid gland of the Cyclostomata. The lateral evagina- 

 tions of the segment-region probably form rudimentary gill-sacs, and 

 may be comjDared with the peribranchial spaces of Tunicates, and the 

 lateral diverticula of the anterior part of the intestine of larvae of 

 AmpMoxus. Behind these is a looj)ed portion ; in the second and 

 fourth of these loops there is communication with the exterior by means 

 of pores ; of the last there are no less than six in Ihe fourth loop. They 

 are probably to be regarded as rudimentary gill-clefts, without valves or 

 skeleton. 



The funnel-shaped organs in the segment have no internal folds, 

 and may apparently be compared with the ectodermal parts of the 

 segmental organs of the first body- segment. Bateson's view that the 

 reproductive organs have some relation to the epidermis is not accepted ; 

 these organs consist of a series of peritoneal outgrowths on either side 

 of the body. Each ovary is attached to a hollow stalk, which represents 

 the neck of the outgrowth, and, like it, consists of modified peritoneal 

 cells. The eggs develope in the walls of the sac, in the midst of meso- 



