1884.] Balanoglossus Kowalevskii (Agassiz), §c. 



27 



to the formation of fibres and occasional ganglion cells from the 

 lowest layer of the skin. 



The Body -cavities. — I have already described the formation of 'the 

 proboscis body-cavity from a single anterior ponch of the archenteron, 

 having two posteriorly directed horns. On the left of the proboscis 

 stalk (one to two gill-slits) an internal thickening of the epiblast 

 arises. A cavity appears in this strnctnre, which becomes open to 

 the exterior. Subsequently it opens internally to the left horn. This 

 opening is the proboscis pore. Spengel states that in B. Kupfferi a 

 similar pore is also present on the right side. In connexion with 

 this fact I may mention that in a single specimen of B. Kowalevskii 

 (seven gill-slits), which was also otherwise abnormal, I found this 

 double arrangement. 



The two collar body-cavities coalesce in the adult on the ventral 

 side, the septum remaining dorsally. Their cavities are prolonged 

 into two horns, which run in the proboscis stalk side by side with the 

 posterior horns of the anterior body-cavities. 



Into the collar body-cavity opens the pair of ciliated funnels leading 

 to the exterior, mentioned by Spengel. He describes them without 

 a figure as arising from the anterior gill -pouch. In B. Kowalevskii 

 they arise as thickenings of the mesoblast, lining the inner fold of 

 the posterior limb of the collar. These thickenings become hollow 

 and strongly ciliated, opening with the first gill-slit into the atrial 

 cavity formed by the collar fold. In B. minutus this atrial cavity is 

 reduced to a minimum ; the posterior skin of the collar runs, there- 

 fore, directly into the skin of the first gill-slit. It appears that these 

 funnels are of an excretory character. This view is supported by the 

 fact that Hatschek has recently mentioned the presence of an excretory 

 tube opening with the first gill-slit of Amphioxus on the left side. 

 The posterior pair of mesoblastic pouches remain separate throughout 

 life, lining the whole body-cavity posterior to the collar. They send 

 forward a pair of horns between the notochord and the neural cord as 

 far as the proboscis pore. 



The first blood-vessel arises as a splitting in the septum formed by 

 the adjacent walls of these two horns, and is continued forwards with 

 their growth. This remains throughout life as the dorsal blood- 

 vessel. The ventral blood-vessel is formed as a split in the ventral 

 septum between the third pair of body-cavities. 



The dorsal vessel is continuous with the cavity which is correctly 

 described by Spengel as immediately overlying the notochord in the 

 proboscis. This cavity is filled with coagulum, its walls are muscular, 

 and it would seem to me to be the true heart. The space overlying 

 it, which is called by Spengel " the heart," contains, as he has 

 described, a variable quantity of loose cellular tissue. This may 

 possibly be of a glandular character ; it is continuous with that of the 



