438 



SCIENCE. 



[N. S. Vol. XXIV. No. 614. 



vague, undetermined, dreams and faith. I 

 offer to collaborate with, say, M. Crepieux- 

 Jamin." 



" Graphology is an art of the future." 

 With slight modification one can agree with 

 Dr. Binet in his conclusions. 



There is unquestionably a trace of the man 

 left in every act he performs, but the trace left 

 in writing has not been shown to be a better 

 guide to a knowledge of the sex of the writer 

 than a footprint; of the age than a view of 

 the garments; of the intelligence than the 

 weight of the brain; nor of the character than 

 the appearance of his umbrella. It is not 

 within the power of true science to say that 

 such and such can never be attained, but so 

 far as graphology is concerned we may cite 

 the experiments of its greatest investigator to 

 prove that as yet it has furnished no reliable 

 means of attaining to a knowledge of sex, age, 

 intelligence or character from handwriting. 



Persifor Frazer. 

 Philadelphia, September, 1906. 



Genera Avium. Edited by P. Wytsman. 

 4to. Brussels: V. Verteneuil and L. Des- 

 met. Part I., Passeres — Fam. Eurylsemidse. 

 By Ernst Hartert. 1905 (1904). Pp. 8; 

 pi. I. Part II., Picarise — Fam. Todidse. 

 By P. Wytsman. 1905. Pp. 4; pi. I. Part 

 III., Psittaci — Fam. Stringopidse. By T. 

 Salvadori. 1905. Pp. 2 ; pi. I. Part IV., 

 Psittaci — Fam. Nestoridse. By T. Salva- 

 dori. 1905. Pp. 3 ; pi. I. Part V., Psittaci 

 — Fam. Cacatuidse. By T. Salvadori. 1905. 

 Pp. 7; pis. II. 



The first five parts of this important work, 

 which is intended to include the birds of the 

 world, are all that have appeared up to the 

 present time. All bear the date 1905, though 

 part one was issued also as a sample niimber 

 during the first half of 1904. Each part is 

 separately paged and contains but a single 

 family. The introductory portion consists of 

 a short historical account of the group treated, 

 its anatomical characters, general habits, 

 range and bibliography. Then follows a key 

 to the subfamilies, if there are any, succeeded 

 by a systematic treatment of the subfamilies. 



genera and species. Under each subfamily 

 there is a key to its genera; while for each 

 genus are given brief synonymy, generic char- 

 acters, geographical distribution, a key to the 

 species and a list of species with geographical 

 distribution and a little, often incomplete, 

 synonymy. 



The Eurylsemidse — more properly Eurylai- 

 midse — (part I.) are divided by Dr. Hartert 

 into two subfamilies — Calyptomeninse and 

 Eurylseminge — the first consisting of a single 

 genus with three species, the second of six 

 genera. One form, Psarisomus dalhousiw 

 horneensis, from the mountains of northwest- 

 ern Borneo, is described as new. The accom- 

 panying plate represents the heads of several 

 species. The general treatment of this group 

 is very satisfactory, but we are not quite 

 sure that all the forms treated as subspecies 

 are not in reality distinct, though, of course, 

 closely allied species. More careful proof- 

 reading, moreover, would have avoided several 

 very unfortunate errors in scientific names. 



The Todidse (part II.), a family restricted 

 to the Greater Antilles, comprise but a single 

 genus, of which four forms are recognized 

 here. Todus pulcherrimu^ Sharpe is treated 

 as a synonym of T. hypochondriacus, and 

 apparently with reason; but we are not at all 

 satisfied that the four admitted forms are 

 merely subspecies, as our author thinks. All 

 are represented on the accompanying plate. 



The New Zealand family Stringopidse — or, 

 as it should be spelled, Strigopidse — (part III.) 

 has only a single genus of two species, one of 

 which is doubtful — probably but an individual 

 aberration. The plate illustrates various de- 

 tails of 8trigops habroptilus, including the 

 head of an interesting xanthochroic variety. 



Of the New Zealand family Nestoridse (part 

 IV.) six species, all in the genus Nestor, are 

 admitted, but two of these are doubtfully 

 valid, and one is extinct. Four of the forms 

 appear on the single plate. 



The Cacatuidse (cockatoos) (part V.) are 

 divided into two sub-families — Cacatuinse and 

 Calopsittacinse. The first is composed of six 

 genera, including provisionally Newton's curi- 

 ous Lophopsittacus from Mauritius, In Ca- 



