Humphrey, Brain of the Snapping Turtle. ioi 



metaplexus. The indications point to the paraphysis being di- 

 rectly concerned in the lymph supply of the brain. 



The relation of these structures in the large sea turtle, 

 Clielone midas, is very similar though it undergoes some modi- 

 fications. The whole dorsal outgrowth (Fig. 7) including epiphy- 

 sis, dorsal sac and paraphysis, is relatively enormous in size. It 

 grows almost perpendicularly to the longitudinal brain axis, not 

 curving forward as it does in CJielydra. The epiphysis is very long 

 and slender with an exceedingly tortuous narrow stalk and is 

 much folded upon itself but has little widening at the distal end. 

 The dorsal sac is very spacious. Its lateral and caudal walls are 

 formed of the single unconvoluted membrane. Unlike the same 

 membrane in Chelydra it is not folded about the epiphysis ex- 

 cept at its distal end. Around the distal end and cephalo-dor- 

 sad of it the membrane is closely convoluted forming what ap- 

 pears to the naked eye a thick, solid roof to the dorsal out- 

 growth. The velum is relatively slightly developed hanging 

 down from the cephalic portion of the dorsal sac but not reach- 

 ing to the brain. The paraphysis is very well marked and is 

 greatly infolded and lobulated. Its cephalic wall drops down be- 

 tween the diencephal and hemispheres and from it are formed 

 the large plexuses of the diencephal and prosencephal. 



Structure of the Thalami. — In figures 6, 7 the supracom- 

 missure is shown as it arches over the diaccele. If we trace 

 this commissure laterad we find that it soon divides into two 

 portions ; the caudal branch curves sharply and descends almost 

 directly ventrad into the body of the thalamus (Fig. 31). The 

 cephalic branch passes laterad and slightly cephalad to the lat- 

 eral margin of the diencephal and then is continued cephalo- 

 ventrad ectad of the optic tracts passing ventrad of the porta 

 and entering the cerebrum immediately laterad of the pedun- 

 cular tract where it turns sharply and is distributed to the ven- 

 tro-occipital portion of the cerebrum. Osborn has found similar 

 relations in CryptobrancJius. Whether the caudal bundle of the 

 supracommissure is the homologue of Herrick's " commissura 

 habenaria " I am not able to say. I am certain however that 

 its fibers do not extend cephalad in their course until they have 



