252 r^P' 



and are, therefore, as likely as not to be tlio home o£ the insect. 

 Examining them in various ways, and more especially watching them 

 in still summer evenings, when the Ephestias are in the habit of 

 flying, may lead to our turning up the insect in greater numbers, and 

 learning something more about it. 



Tarrington, Ledbury : 



February, 1888. 



NOTES ON DR. HERMANN MCLLER'S "FERTILISATION OF 



FLOWERS." 



BY EDWABD SAUKDEES, F.L.S. 



I have recently had occasion to consult the late Dr. Hermann 

 Miiller's " Fertilisation of Flowers," as translated by Mr. D'Arcy W. 

 Thompson, but although I have been, as every one must be, much 

 interested in his remarks on the ILymenojptera given at p. 45, et seq., 

 I feel compelled to make a few observations on some of the data on 

 which he founded his deductions, and which appear to me to be 

 incorrect, and, therefore, likely to cause such deductions to be mis- 

 leading. I do not discuss Dr. Miiller's theories as to the ancestry of 

 the Anthophila, as I feel myself quite incompetent to do so ; hut my 

 first remarks shall be on the genus which he considered first among 

 the bees, viz., Prosopis. 



Of the species of this genus, he says, p. 47 : " In their almost hair- 

 less bodies, the narrow first tarsal joint, scantily provided with hairs, 

 and their very slightly elongated mouth parts^ they completely 

 resemble the sand wasps, and only claim to be admitted to the family 

 of bees by their manner of feeding the young.", 



Dr. Miiller apparently overlooked the beautifully pectinate hairs 

 on the thorax and abdominal bands of the various species of this 

 genus, and also the branched hairs which exist on the coxae, tro- 

 chanters, and femora (I have figured a metathoracic hair of Prosopis : 

 Trans. Ent. Soc. Lond., 1878, pi. vi, fig. 14), affording characters which, 

 if my observations be correct, distinguish them at once from any of 

 the family of sand wasps. From Prosopis Dr. Miiller goes at once on 

 to SpTiecodes, so that he altogether omits Colletes, another genus of 

 the Ohtusilinguce, which has the tongue short and bifid, as in Prosopis, 

 but which is densely clothed in all the known species with beautifully 

 branched hairs, in fact, the hairs of Colletes and its kindred Austra- 

 lasian genera Lamprocolletes, &c.,are probably more strikingly branched 

 than those of any other known genera (for figures of hairs from 



