ZOOLOGY AND BOTANY, MICEOSCOPY, ETC. 373 



under existing conditions, and from pseudo-relicts which migrated when 

 the area was accessible in a way in which it no longer is. A glacial 

 relict must be a stenothermal inhabitant of cold water, not cosmopolitan 

 in distribution, and it must occur in an area where the previous 

 occurrence of glacial conditions is demonstrable. 



Relations of Parasite's Phylogeny to that of Host.* — Launcelot 

 Harrison maintains that in the case of total obligate parasites, closely 

 related forms will be found on phyletically related hosts, without regard 

 to other cecological conditions. 



Like Kellogg, he bases his case chiefly on Mallophaga, but there is 

 other evidence. The Auopluran genus Pediculus is represented on man, 

 on Simiidae, and on Ateles alone among Cebidse ; the closely related 

 genus Fedirinus occurs on lower monkeys. Similarly, the parasitic 

 Platy helminths of a given type have their nearest relatives in a host 

 phyletically related to theirs. 



Many Mallophaga have a world-wide distribution, but always on 

 closely related birds. Thus Philopterus lari occurs on all gulls ; 

 Lvpeurus anatis on all ducks ; L. columhse on all pigeons ; Golpocephalimi 

 Jiavescens on all hawks. The genus Tetrophthalmus lives in the gular 

 pouch of pelicans, and has its tracheal system specially modified in 

 relation to its peculiar hal^itat. Species of Tetrophthalmus occur on all 

 pelicans, in the same situation, and are all similarly modified. The only 

 reasonable explanation is that the parasites have had a common origin. 



In their equable conditions the Mallophaga have not evolved as 

 rapidly as their hosts. The Philopterus of the European cuckoo is with 

 difficulty separable from species from the Australian genera Cacomantis 

 and Ghcdcococcyx. 



INVERTEBRATA. 



Arthropoda. 



a. Insecta. 



Germ Cells of Hymenoptera.t — R. W. Hegner has made a series of 

 studies on protoplasmic differentiation in the oocytes of Hymenoptera. 

 The first deals with the differentiation of the oocytes and nurse cells in 

 the ovaries of the hive-bee. Four regions may be recognized in the 

 ovariole of the queen honey-bee : (1) the terminal filament ; (2) a 

 rosette region ; (3) a zone of differentiation ; and (4) the posterior part, 

 in which the oocytes form a linear series separated from each other hj 

 groups of nurse cells. The rosette region is filled with rosette-like 

 groups of cells, each group consisting of the descendants of a single 

 mother oogonium. The cells of a rosette are united by strands which 

 are the persisting spindle fibres from earlier mitoses. The cells in a 

 rosette divide synchronously. Oocytes and nurse cells are both derived 



* Rep. Brit. Assoc. Manchester, 1916, pp. 476-7. 



t Journ. Morphology, xxvi. (1915) pp. 495-560 (13 pi.). 



