82 BRACHIOPODA. 



first segments become abstracted from the anterior part of the body. 

 A consideration brought forward by Caldwell seems of still greater 

 importance. It is evident that in Adinotrocha the principal axis of 

 the larva differs from that of the adult. In the larva, this axis runs 

 from the neural plate to the anal aperture, and corresponds to that 

 of the Trochophore and of the Annelid that develops from the 

 IVochophore. In Phoronis, after metamorphosis, on the contrary, a 

 secondary principal axis is met with at right angles to the primary 

 axis of the Actinotrocha. The principal axis of the Brachiopodan 

 larva, however, corresponds to the secondary axis of Phoronis^ and 

 runs from the neural plate to the point of attachment, which must 

 be imagined as lying in the middle of the ventral side proper. 

 When, therefore, segmentation takes place in a direction transverse 

 to this principal axis, this cannot be compared to the segmentation 

 of the Annelida, because the position of the latter is determined by 

 the direction of another principal axis (the primary axis of an 

 Actinotrocha). This example shows that the comparison with the 

 Actinotrocha is of great utility in forming conclusions as to the 

 Brachiopoda. 



If we believe in the comparison of the Brachiopoda with Phoronis 

 and the Bryozoa, in which two groups the short line lying between 

 the mouth and the anus is to be recognised as the dorsal middle line, 

 we shall be in agreement with Caldwell, who considers that the 

 two shell- valves in the Brachiopoda strictly speaking belong to the 

 ventral side, and that here also attachment takes place at the middle 

 of the ventral side. 



We have seen that in the Brachiopodan larva there is no true 

 segmentation, and this is also the case with the adult. The presence 

 of two pairs of segmental organs in Rhynchonella is the strongest 

 evidence that can be brought forward in favour of segmentation. 

 Yet we know that in the Annelida more than one pair of segmental 

 organs may occur in a segment, so that too great stress should not be 

 laid on this feature alone. In the same way two pairs of nephridia 

 also appear in Phoronis australis (Benham).* 



There are many reasons for believing that the Ecardines must be 

 regarded as the most primitive forms of the Brachiopoda. In this 

 division a lateral anal aperture is retained in Lingula, and one lying 

 posteriorly in the median plane in Crania. The peduncle in Lingula 



• [While pointing out that each of the two nephridia has a double opening 

 into tin- body-cavity, BENHAM never commits himself to the assertion that there 

 •ii. two pain <•!' nephridia. He considers Phoronis as more closely allied to the 

 Gephyrea than to the Bryozoa.— Ed.] 



