March 11, 1904.] 



SCIENCE. 



425 



development and, by complete absorption, 

 leave no trace in the fully mature grain. 

 For the determination of this point the fol- 

 lowing forms were examined: Cupressus 

 (four species), Taxus baccata and four va- 

 rieties, Jtcniperus (two species), Chamce- 

 cyparis (five species), Callitris (one spe- 

 cies), Cryptomeria japonica, Thuja orient- 

 alis. 



The results showed that in all these forms 

 there was no prothallial cell formed at any 

 time in the development of the pollen. 

 Ovules of Thuja orientalis and Taxus iac- 

 cata were examined to determine the num- 

 ber of potential megaspores formed. It 

 was found that in Taxus . there are four 

 produced— these lying, as a rule, in a row, 

 and the lower developing into the pro- 

 thallium. 



In Thuja there are also four magaspores 

 produced, but they are arranged, not in a 

 row, but in nearly regular tetrad form, dif- 

 fering in this respect from all other gym- 

 nosperms so far studied. 



W. F. Ganong, 



Secretary. 

 Smith College, 

 Noetuampton, Mass. 



SCIENTIFIC BOOKS. 



STILL ANOTHER MEMOIR ON PAL^OSPONDYLUS. 



Palmospondylus, like Gloster, seems to have 

 been born to bite the world — for in its few 

 short years of morphological nurture it has 

 succeeded in causing trouble to an amazing 

 degree. And we venture to use an Elizabethan 

 simile with a clearer conscience, since in the 

 latest time we are told that this obscure little 

 fossil is not to be looked upon as a toothless 

 lamprey, but rather as a shark. We must 

 none the less admit that it gives us a sense of 

 sadness to learn of the new role of the fish, 

 since this reduces by one the novelty of its 

 being assigned to still other groups, for we re- 

 call that the number of groups is well nigh ex- 

 hausted. There remains in fact but one more 

 of the major groups of aquatic vertebrates to 

 which it can possibly be assigned. It has 



already been reckoned among arthrodires 

 (? ostracophores), lampreys, teleostomes {Al- 

 lis), sharks, lung-fishes, even amphibians. 

 To make our list complete, we have now only 

 to assign it to holocephals. And lest some 

 one anticipate us, we may as well regard it as 

 a chimseroid at this time, and in evidence 

 of this refer to its continuous dorsal fin, 

 'protocercal ' tail, ring vertebrse, elaborate 

 nasal cartilages, huge head — and we might 

 find other similarities if we tried hard enough. 

 Seriously, though, such a state of affairs is a 

 reproach to modern morphology, that with all 

 our extensive knowledge of fishes we are not 

 able to come to a better understanding of our 

 Devonian ' lamprey.' For if the remains of 

 Palmospondylus are so poorly preserved that 

 they cannot be definitely described, why do 

 we continue to add papers to the troublesome 

 literature? The only possible excuse is that 

 the creature is seductive, full of suggestions 

 as to the origin of the gnathostomes, and the 

 mode of evolution of jawless vertebrates. 



During the past summer I happened to see 

 in South Kensington some of the elaborate 

 models of Palwospondylus which Professor 

 Sollas and Ingerna B. J. Sollas have been 

 preparing. These are built up of thin wax 

 plates, after the method of Born, the sections, 

 however, having been outlined at a series of 

 levels (differing in thickness for about .1 +, 

 .025 mm.) as the fossil was carefully ground 

 down. And I examined the models with great 

 interest, wondering whether by a new method 

 there could be gained facts which would help 

 to solve the present puzzle. 



In their complete paper* the authors now 

 discuss the results obtained from a series of 

 their models, of which no less than eight have 

 been prepared. They describe the character- 

 istic parts of the fossil, christening some of 

 them with rather difficult names such as 

 ' hemidome,' ' tauidion,' ' ampyx,' ' gammation ' 

 and ' pre-gammation.' But the structures 

 described which interest us especially are the 

 ' branchial arches,' four in number, showing 



* ' An Account of the Devonian Fish, Palwo- 

 spondylus gunni, Traquair,' by W. J. Sollas and 

 [ngerna B. J. Sollas, Phil. Trans., Series B, Vol. 

 196, pp. 267-294, pis. 16 and 17. 



