NOTBMBEE 27, 1914] 



SCIENCE 



111 



beyond this point the percentages of eggs fertil- 

 ized should fall off at first slowly, then rapidly, 

 and then slowly again to a vanishing point. 



This condition is not realized, however, in actual 

 experiments unless the time interval between pre- 

 paration of the more dilute sperm suspensions and 

 addition of the eggs is made very short (less than 

 five minutes). Under such optimum conditions 

 the curve of percentages of fertilized eggs begins 

 to fall from 100 per cent, at a dilution of about 

 1/3000 of 1 per cent, sperm; the curve falls slowly 

 to about 1/24000 per cent., then rapidly to about 

 1/300,000 per cent., then slowly again to about 

 1/90,000,000 per cent, where however about 1 per 

 cent, of fertilizations may still taie place. The 

 observations show that beyond a dilution of 

 1/20,000 of 1 per cent, only a single spermato- 

 zoon can possibly be concerned in the fertiliza- 

 tion of each egg. 



If the time interval be lengthened to twenty 

 minutes, fertilizing power of sperm suspensions 

 may be completely lost at 1/1000 per cent., a 

 point in the series of dilutions at which each egg 

 recieves several spermatozoons. Comparing sperm 

 suspensions of increasing dilution it is found that 

 the rate of loss of fertilizing power is inversely 

 proportional to concentration. Thus the time re- 

 quired for complete loss in a series of sperm sus- 

 pensions between 1/600 per cent, and 1/120,000 

 per cent, forms a curve ranging from 45 down to 

 6 minutes. Presumably the time is even shorter 

 at greater dilutions. 



Spermatozoa may be perfectly motile after loss 

 of fertilizing power. Their ineffectiveness in 

 these experiments is therefore due to loss of a 

 necessary substance. This is an interesting con- 

 firmation of the postulate, for which all experi- 

 mental proof has hitherto been lacking, that the 

 fertilizing power of spermatozoa is due to a 

 definite substance. The spermatic substance in 

 question represents the ' ' sperm-receptors ' ' of my 

 theory of fertilization. 



W. L. Towee: Experimental Production of a New 



Character. 



The antennae of the Chrysomelidffi are highly 

 invariable organs and are used but little in taxo- 

 nomic differentiation, even of the genera and 

 families. Of especial interest, therefore is the 

 experimental production, by means of continued 

 environmental pressure of a factorial group that 

 is productive of antennal conditions not known to 

 exist in any living or fossil Chrysomelidas. In 

 origin it arose progressively, in one direction. 



exists in three states of stability, each of which 

 is capable of transference to other species through 

 crossing, thus giving a picture of what may be 

 one method of the production of nearly related 

 genera. A final point of significance is that its 

 behavior in crosses is no criterion of its method 

 of origination, as it arose progressively, with all 

 possible intergrades, but was at all points in the 

 series alternative and dominant to the normal. 



S. W. "Williston: The American Land Vertebrate 



Fauna and its Melations. 



The land vertebrate fauna of Lower Permian 

 or Permocarboniferous age in North America 

 comprises, so far as now known, at least sixty 

 definitely distinct genera, distributed about equally 

 among the Amphibia, Cotylosauria and so-called 

 Pheromorpha. From all other parts of the world, 

 of approximately equivalent age, less than a dozen 

 genera are known, for the most part imperfectly. 

 In North America vertebrates are known only 

 from New Mexico, Texas and Oklahoma, Illinois, 

 and Pennsylvania. The fauna of New Mexico 

 comprises, so far as is yet known, sixteen valid 

 genera, twelve of them unknown elsewhere, the 

 remaining four somewhat doubtfully identified 

 with Texas forms. Not a single genus or family 

 even of the American fauna is definitely known to 

 occur elsewhere. 



The American forms, and especially the higher 

 reptiles, within the limits of their more general- 

 ized characters, are very diverse. That localities 

 so little remote as Texas and New Mexico, though 

 showing intimate family resemblances, should be 

 so distinct in their genera is evidence that the 

 world's fauna in Lower Permian times was an 

 exceedingly abundant one. Probably at no time 

 in the earth's history has there been a more ex- 

 tensive fauna of reptiles. As it is, there is no 

 formation known in geological history of ap- 

 proximately equal duration that has yielded a 

 greater number of genera of reptiles and amphi- 

 bians than the American deposits. 



The conclusion is legitimate that as early as 

 the close of Carboniferous times the reptilian 

 fauna of the world was a relatively old one. It 

 has been urged that the relationships of this 

 reptilian fauna with that of the Middle and 

 Upper Permian of Africa was a close genetic 

 one. From a recent study of most of the known 

 specimens of European Permian vertebrates I am 

 convinced that their resemblances are yet closer. 

 On the other hand it has been urged that the 

 American Permian fauna is an isolated one, with- 



