228 MUTUAL EELATIONS OF COLD-BLOODED TEKTEBEATA. 



9^ 



' Exo-skeleton witli bony scutes, or with roiigli scales or 

 spines 10. 



Exo-skeleton, in most, with ctenoid or cycloid scales. Bran- 

 chiae pectinate 11. 



r Branchiae clustered. Body elongate. . Orel. 1. Lophobranchii. 



10 \ Branchiaj pectinate. Body short. Maxillaries connate mth 



[ pre-maxillaries Orel. 2. Plectognathi. 



{Swimmiug-bladder with air-duct. All the fin-rays, except some- 

 times the first, soft, jointed . . Orel. 3. 'Phtsostomi. . 12. 

 Swimming-bladder without air- duct 13. 



, o J Ventral fins absent r Suiord. 1. P. Apodes. 



I Yentral fins on abdomen Subord. 2. P. Abdomiuales. 



, o f Lower hyoid bones united. . . . Ord. 4. Phaetngognathi 14. 

 \ Lower hyoid bones distinct 15, 



(All the fin-rays, except sometimes the first, soft, jointed. 

 Subord. 1. P. Malacopterygii. 

 Anterior fin-rays hard, unjointed. 

 Subord. 2. P. Acauthopterygii. 



I Pin-rays soft, jointed. Ventrals beneath pectorals, or absent. 



^ _ J Ord. 5. Anacanthini. 



I Anterior fin-rajs hard, unjointed. Ventrals usually beneath, 



(^ or before, pectorals Ord. 6. Acanthopteei. 



In the above arrangement, the relative positions of the orders 

 as indicated by Miiller have been carefully maintained *. 



* Miiller establishes a sixth subclass {Dipnoi) for the Lepidosiren. A 

 further contribution to our knowledge of this genus has just been made by Dr. 

 E. M'DonneU (see Natural History Eeview, April, 1860). The large size of 

 the blood-disks hi Lej^idosiren, the general arrangement of its circulatory 

 apparatus, audits power of emitting vocal sounds are in this memoh referred to 

 among other proofs of its near affinity to the Ampliibia. Dr. M'Donuell cou- 

 fii'ms the biaui'icular character of the heart and the presence of posterior nares. 

 These observations were based on the examination of eight specimens of L. an- 

 nectens. To this paper a note has been appended by Professor Melville, who 

 strongly contends for the Ampliibian natui-e of the genus, which he proposes 

 to place between the Ferennibranchiaia and Derotremata of Stannius. 



