906 



SCIENCE 



[N. S. Vol. XXXI. No. 806 



somes may therefore be classed as a diploid 

 growth in distinction to the original gameto- 

 phytic or haploid stage with its In number 

 of chromosomes. These diploid growths of 

 dioecious (heterothallie) species remain en- 

 tirely sterile, though producing apparently 

 normal antheridia and archegonia. Attempts 

 made to produce hybridization between these 

 hermaphroditic diploid growths and male and 

 female plants of the normal In generation 

 resulted in failure. 



With hermaphroditic (homothallic) species 

 the condition is different and the aposporic 

 outgrowths are fertile. Their gametes with 

 2n chromosomes unite and produce sporo- 

 phytes with 4/1 chromosomes. These tetra- 

 ploid sporophytes form spores with again 2ra 

 chromosomes, which grow into fertile gameto- 

 phytes with double the normal chromosome 

 number, thus producing a definitely fixed bi- 

 valent race (e. g., Amhlystegium serpens hi- 

 valens). The regeneration from the tetra- 

 ploid sporophyte gives rise to a race with 4?i 

 chromosomes which as yet has remained ster- 

 ile. A sporophyte with 8n chromosomes 

 raight be produced if this 4n race could be 

 induced to fruit. 



No phenomena have been observed, such as 

 apogamy or supplementary chromosome reduc- 

 tion, which would avoid the doubling of chro- 

 mosomes in the races obtained from sporo- 

 phytic regenerations. 



A rather careful series of measurements 

 were made of the size of the nuclei and cells 

 in the different stages obtained, and it was 

 found that the volume of the cells and of the 

 nuclei were directly proportional to the num- 

 ber of chromosomes in the In, 2n and 4?i 

 gametophytes as well as in the 2n and 4n 

 sporophytes. Further, it was seen that this 

 increase in size of the cells with an increase 

 in the number of chromosomes resulted in the 

 enlargement of certain organs such as the 

 antheridia and archegonia. 



The Marchals believe that apospory is likely 

 to occur in nature from wounding of the 

 sporophyte and that bivalent races have thus 

 been formed. 



There is promised a continuation of these 

 investigations on the mosses which have 



proved already of such great interest to the 

 students of sex. 



A. F. Blakeslee 

 Connecticut Agbicultubal College, 

 Stobbs, Conn. 



A List of Geographical Atlases in the Library 

 of Congress, with bibliographical notes. 

 Compiled under the direction of Philip Lee 

 Phillips, F.E.G.S., Chief Division of Maps 

 and Charts. In two volumes, cloth: Vol. 

 I., Atlases, pp. xiv + 1,208, Vol. II., Author 

 List, Index, pp. 1,209-1,659. "Washington, 

 Government Printing Office. 1909. $2.35. 

 In the publication of these volumes a very 

 commendable service has been done for geog- 

 raphy and for students in all lines making 

 use of maps. For it is strictly true as the 

 editor says in his preface, " atlases have not 

 received the consideration in bibliography due 

 to their importance in literature and as con- 

 tributions to knowledge." There are few 

 works on the subject and these are fragmen- 

 tary. 



The present contribution is merely a list 

 of the geographical atlases in the Library of 

 Congress, a total of over thirty-four hundred 

 titles in addition to seventy lettered titles. 

 The editor modestly disclaims it as a bibliog- 

 raphy. 



The arrangement is good. It starts with 

 general atlases of special subjects, the subject 

 headings in alphabetical order. Then follow 

 the general atlases in chronological order, and 

 then follow America, Europe, Asia, Africa 

 and Oceanica in similar order. This classifi- 

 cation includes under each general heading 

 the atlases of cities, of voyages of circum- 

 navigation, historical works, scientific ex- 

 plorations, and the atlas material accompany- 

 ing the reports on boundary disputes between 

 nations. 



Bibliographical notes and tables of con- 

 tents have been given in case of the rare and 

 more important volumes. This brings out 

 numerous inserted maps, so frequently hidden 

 away in such material. 



In the second volume the general index is 

 preceded by an author index of abridged 

 titles, in which the author's full name is 



