March 8, 1906] 



NA TURE 



439 



London. In the libraries of the learned societies at 

 Burlington House alone there are many serials in dupli- 

 cate ; some of these might profitably be replaced by others 

 which are not at present in these libraries. It often 

 happens that books and serials are sent to library com- 

 mittees on approval, and are rejected because they are 

 thought to be more suitable for other libraries ; but attempts 

 are not always made to ascertain whether these other 

 libraries possess them. At the present time, from want of 

 space and other causes, the duplication of periodicals at 

 Burlington House is avoided as much as possible. 



In the subject index to the scientific literature of the last 

 century which the Royal Society is preparing, it is pro- 

 posed to indicate, in the introductory list of serials, the 

 libraries in which the cataloguing has been done, and also 

 to mention other libraries in which the books may be found. 

 This will be useful to workers, but it cannot be quite 

 complete, for the task would be too great to undertake in 

 its entirety. For example, there are more than 600 serials 

 which contain mathematical papers, and it would be im- 

 possible to name all the libraries where they are found. 



March 3. Herbert McLeod. 



The Bees of Australia. 

 Up to the beginning of 1905, 224 species of wild bees 

 had been recorded from Australia, no less than 183 of 

 them having been described by F. Smith, of the British 

 Museum. I had the opportunity in 1904 to study Smith's 

 types at the British Museum, and since then I have worked 

 up the unnamed Australian material belonging to that 

 institution, with the exception of some species of Halictus 

 yet to be examined. The following list shows the genera 

 found in Australia (including Tasmania), New Zealand, and 

 the Austro-Malay Islands (taking the region as defined by 

 Wallace), and the number of species in each. 



Family. 



Genus. 



Australi; 



New Austro- 

 Zealand. Malay 

 Islands 



— 21 



no. 1897, VOL. 72s\ 



317 



The list proceeds from the most primitive bees up to 

 the most specialised. The genera marked with an asterisk 

 are wholly peculiar to Australia, so far as known ; and 

 it will be' observed that, as with the mammals, there are 

 many endemic genera of a primitive type. Lestis is the 

 only endemic genus allied to the ordinary long-tongued 

 bees, and that consists of two closely allied species, which 

 represent an offshoot from Xylocopa, probably not of very 

 ancient date. True Xylocopa, it will be noticed, just enters 

 Australia (but one species is common in the north), but is 

 rich in species in the Austro-Malav Islands, and extends 

 into Asia, Europe, Africa, and America. The Xylocopas 

 are the large carpenter bees, which nest in wood, and may 

 be transported across the water in floating trees. Until 

 recently, the genera Thaumatosoma and Exoneura were 

 supposed to be peculiar to Australia, but the first has now 

 been found in Burma and the second in Syria. _ They may 

 possibly be genera which are verging on extinction, but 

 as each differs only in one important particular from its 

 nearest ally (these allies being Megachile and Allodape 

 respectively), it is not impossible that they arose by parallel 

 mutations in the widely distant localities in which they 

 occur, quite independently. 



The most interesting of the primitive genera is Phena- 

 colletes, based on a new species (P. mimus) discovered by 

 Commander J. J. Walker on the Penguin Expedition. The 

 Colletid bees are supposed to have been derived from the 

 fossorial wasps, and Phenacolletes is so like certain wasps 

 that I was not sure whether it was a wasp or a bee until 

 I had examined its pubescence with a compound micro- 

 scope. Unfortunately, we know nothing of the habits of 

 this insect, but Commander Walker kindly informs me 

 that it was taken on November 12, 1890, at Turtle Bay, 

 north end of Dirk Hartog Island. He finds in his journal 

 for that day that " an upright growing shrub with ovate 

 glabrous leaves and large whitish-rosy mallow-like 

 flowers " was the only plant which seemed to be at all 

 attractive to insects, so perhaps the Phenacolletes came 

 off that. 



I have supposed that the bees with emarginate tongues 

 (Colletids and Prosopids) arose from the wasps indepen- 

 dently from those with pointed tongues, this seeming the 

 more likely, because the wasps themselves exhibit both 

 types. However, there are indications that in Australia 

 the first form may have become modified into the second 

 within the limits of the bee-group. This is especially 

 suggested by the tongue of Callomelitta, and by one of the 

 new species' placed for the present in Paracolletes. 



The new genus allied to Callomelitta, indicated in the 

 table, is for Sphecodes antipodes, Smith. Colonel Bingham 

 very kindly made a critical examination of this species at 

 my request, and found that it was not a Sphecodes, but 

 belonged to a new genus differing from Callomelitta in the 

 shape of the thorax, pubescence of hind tibiae, &c. It 

 will undoubtedly prove an important form from the stand- 

 point of the evolutionist. 



The species marked as Andrena? and Scrapter? stand 

 in our lists as members of these northern genera, but they 

 have not been critically examined recently, and it is 

 questionable whether they are rightly classified. The name 

 Mellitidia has been applied to the so-called Andrena of 

 Australia, and it is probably valid. Nevertheless, there 

 are some undoubted cases of well known northern genera 

 having endemic Australian species, while they have none, 

 so far as known, in the Austro-Malay region. These are 

 Nomioides (found from Burma to Europe) and Tetralonia 

 (India to Europe, &c.) ; Saropoda (also European) is really 

 in the same category, as the single Austro-Malay species 

 is one of the Australian ones, which has reached the Aru 

 Islands. The case of the Tetralonia seemed a little 

 doubtful, but Colonel Bingham has critically examined 

 Smith's type, and reports that it is a true Tetralonia, but 

 is a female, not a male, as Smith had it. Lithurgus is 

 also a genus of Europe and Asia, and likewise Africa, 

 which has Australian species, though none are known 

 from the Austro-Malay islands. In this case, it is prac- 

 tically certain that the genus is dispersed more or less 

 through the islands, and has been overlooked, for one of 

 the Australian species is exceedingly close to one of India. 



Gastropsis, placed by Ashmead in the Andrenidae, is 

 apparently allied to the European Meliturga, and is in a 



