4 NEMERTINES OF PLYMOUTH SOUND. 
Il. MESONEMERTINI (Cephalothriz and Carinoma).—Uateral 
nerves have penetrated the circular muscular layer, and lie embedded 
in the longitudinal layer. 
Il. METANEMERTINI (Hoptonemertint of Hubrecht). — 
Lateral nerves have penetrated the longitudinal muscular layer, and 
lie in the body parenchyma. 
IV. HETERONEMERTINT (Scuizonemerrini of Hubrecht, to- 
eether with the genera Hupolia and Valencinia).—Lateral nerves in 
the same position as in Carinella, but between the epidermis and the 
circular muscular layer a layer has developed consisting of gland- 
cells, connective tissue, and longitudinal muscles, in which the nerve- 
stems hie. 
Of these divisions, that of the Protonemertini is of course regarded 
as the oldest, and Birger regards it as giving rise to the Mesonemertint 
through Carimella, and to the Heteronemertinit through his new form 
Hubrechtia desiderata, for which he has established a new family, 
Hubrechtiade. The Metanemertini he considers to have arisen from 
the Mesonemertint. 
The subordinate changes concern the order Heteronemertini. This 
order is divided into two families,—the Hupoliadx, including the 
genera Hupolia and Valencinia ; and the Inneide. The latter 
consists of two sections, the Micrurx# and the Amicrure, characterised 
by the presence or absence of a caudal appendage. To the 
Amicrurx belong the genera Micrura, Cerebratulus, and Langia, Of 
these the first two were united by Hubrecht under the name of the 
second. They are now again separated, the differentiating characters 
being those of general shape and mode of life. 
In the above arrangement of the Lineide there is obviously a 
return to the scheme laid down by McIntosh. Hubrecht objected to 
the caudal appendage being regarded as a feature of generic value, 
and while thus abolishing the genus Micrura and including its 
species under Cerebratulus, he added to the latter all the shorter and 
broader species included by McIntosh under Linews, with the result 
of establishing two genera, Lineus and Cerebratulus, with absolutely 
no point of difference except a very problematical ontogenetic 
difference. With our present knowledge there are, it appears to 
me, only two ways out of this difficulty. Hither the genera Lineus 
and Cerebratulus must be fused into one, with the result of forming 
a very large genus—a method of escape suggested by Hubrecht 
himself (8) ; or there must be a reversion to the McIntoshian system, 
which, with some alterations in detail, is the method adopted by 
Burger. Of the two alternatives there can be no doubt that the last 
