Dec. 28, 1876] 



NATURE 



179 



on the same footing with the hunling-wasp, and, as to 

 the length of its tongue, is even but little more highly 

 developed. The assiduity with which it visits the flowers, 

 its practice in seeking those where the honey is con- 

 cealed, its well-established habit of feeding its young 

 entirely on food derived from flowers, all place it above 

 the average of its ancestral species. Even, however, 

 among these ancestral species — the hunting-wasps— are 

 certain kinds whose love for flowers has far surpassed their 

 power of rearing a family, and these, favoured by their 

 large size and length of tongue, have attained still greater 

 dexterity in getting at concealed honey than Prosopis. 



Out of eighty-five visits to flowers by different species, 

 there were of Prosopis : — 



On flowers with the honey apparent, 19 = 22*4 per cent. 



On flowers with the honey somewhat concealed, but 

 still fairly apparent, 19 = 22*4 per cent. 



On flowers with the honey concealed, but easily reached 

 through the formation of the mouth, 28 = 32"9 per cent. 



On flowers with the honey entirely concealed, which 

 are specially adapted to the burrowing Hymenoptera, 

 8 = 94 per cent. 



On flowers without honey, or at least only having pollen 

 as far as Prosopis is concerned, 11 = I2'9 per cent. 



Some of the hunting-wasps appear to show even more 

 intelligence than Prosopis in the search for honey, and 

 this will appear still more forcibly if we leave out the 

 flowers which only yield pollenj when the total result is for 

 the wasps 35 "3 per cent., for the Prosopis 48'6 per cent. 



The following table will show at a glance how far the 

 hunting-wasps are above the leaf-cutting wasps, how far 

 the former are surpassed by Prosopis, and how far the 

 hunting-wasps, most eager for flowers, again surpass the 

 Prosopis, 



a b a-\-b c d c+d e J 



Leaf-cutting wasps 69'! i6'3 85*4 i3'o o 130 o id 



Majority of wasps indifferent! g _ 



to flowers ) jj J ' »-> ' -' ■> 



Principal huntinjf wasps ... 44"7 190 637 30 "o 4"8 34'8 I's — 



Prosopis 22'4 22 4 44 '8 32'9 94 42"3 la'g — 



The wasps most eager for) .0 .„ ,„.„ ,e. cjl 



flowers.. f ^° 231 33 9 492 169 661 — — 



If we look down the perpendicular lines of this tabic 

 from top to bottom we shall see the frequency of the 

 visits to the flowers with the honey apparent {a) even if 

 we add the visits to the flowers with the partially seen 

 honey {a-\-b) steadily decline, though the visits to the 

 flowers with partially concealed honey {b) steadily in- 

 crease. In still more rapid ratio the frequency of the 

 visits to the honey entirely concealed, increases {c -\- d), 

 both in the case of those flowers whose honey is easily 

 attainable through the simple form of the corolla (it), and 

 those specially adapted to the burrowing Hymenoptera {d). 



With regard to the flowers yielding only pollen (^), the 

 Prosopis shows a remarkable advance on the other 

 Hymenoptera, probably caused by the great need the 

 young fed from flowers have of the pollen. 



The results of a careful series of observations, pro- 

 longed over a course of five summers, brings us to the 

 following conclusions : — 



1, The Tenthredos stand, as regards the rearing of 

 their young and in their general intelligence, as well as 

 in their capacity of finding and acquiring concealed honey, 

 the lowest of all the Hymenoptera we have considered. 

 They visit flowers with the honey altogether apparent 

 much more frequently than those with the honey wholly or 

 partially concealed, and the latter only in the case of 

 those with a low and simple organisation. 



2. The hunting-wasps, which, as has been previsusly 

 shown, stand decidedly above the Tenthredos, show also 

 a decidedly higher power of finding honey. Even when 

 the most intelligent and flower-loving families are ex- 

 cluded, a high percentage among the hunting-wasps is 

 found visiting flowers with concealed honey, and a lower 

 visiting flowers where the honey is apparent ; they even 

 seek some flowers whose formation shows them to be 

 adapted to the fossorial Hymenopterji. All this progress 



appears still nure rein.'ukablc when wc compare the whole 

 group of Tenthredos with thewhole group of hunting-wasps. 



3. Prosopis, the least highly-developed bee, and one 

 which does not stand higher than the hunting-wasps, 

 cither in general capacity or in care for its young, is yet 

 compelled, through the greater requirements of the latter 

 and their need of pollen, to visit both flowers with con- 

 cealed honey and flowers which only produce pollen, 



4. Certain species of hunting-wasps stand above the 

 Prosopis, both in size and in the length of the proboscis ; 

 these, however, only seek food from flowers for them- 

 selves, and not for their young. 



We may now hope through the following steps to ob- 

 tain some idea how the honey-bee has acquired its 

 peculiar capacity for finding concealed honey. 



All our European bees, except Prosopis and the para- 

 sitic species, have their hairy covering specially adapted 

 to collect pollen. This apparatus forms a characteristic 

 by which they are divided into two families, namely, 

 those which collect the pollen wiih a brush on the lower 

 side of the body, and those which collect it with the 

 pencil-like hairs of the hind legs. 



In the former family there are various degrees of 

 development in the pollen-collecting apparatus, difficult 

 alike to describe or distinguish ; and as the relationship 

 of these species to the honey-bee is but remote, we may 

 content ourselves with glancing at this group as a whole. 



The second group, on the contrary, offers a steady series 

 of degrees of development which we may now compare 

 with each other and with Prosopis. 



Passing over the first step — sphecodes, we come to 

 Andrena and Halictus, whose size and length of pro- 

 boscis places them distinctly above Prosopis, Their visits 

 to flowers rise from 7*4 per cent, in the case of flowers 

 with the honey altogether apparent, to 5r6 per cent, to 

 the flowers of complex form and with the honey entirely 

 concealed. 



The genera Eucera, Anthophora, and Sarapoda reach 

 the highest point of intelligence as regards flowers of all 

 our native bees, the length of their proboscis varying from 

 ten to 21 mm., one of the humble-bees {Bo/nbus horforuin) . 

 being the only one that can in this respect compare with 

 them. 



The three above-mentioned genera, out of fifty-six visits 

 to flowers, paid none to those with the honey apparent, 

 but 91 •! per cent, to those with either concealed honey of 

 complex form, or with both characteristics united. 



There is no further example of increased development 

 of the proboscis in this family, but they have attained a 

 further step in the transformation of the thigh brushes 

 into pollen-baskets, and above all in the establishing the 

 comtnunity of life and of labour which distinguishes the 

 humble hive bees. From observations made on the lat- 

 ter, however, it appears that increased development shows 

 a certain return to the flowers containing more visible 

 honey, the honey bee, out of 520 visits observed, going to 

 such flowers at the rate of o'8 per cent., the hive-bee, 194 

 visits being observed, at the rate of 3'i per cent. 



The following table is fuller than the previous one :— 



a b a + b c d c + d e / g 



Tenthredos 691 i6'3 854 130 o 13-0 o — 16 



Hunting wasps 447 190 637 300 48 348 i'5 — — 



SpL'^".. ::: :::}"-^ -^ ^-^-^ 33-3- r^ .o-^.r,.-, o 



Andrena and Halictus 7^4 20-5 339 402 114 5i"6 96 a'o 2*9 

 CUissa '\ 



Panurgus, Dasypoda 



Rhophites | 



Halicloides J 



Eucera Anthophora ...I 



Saropoda I 



Osmia Chalciodoma...! 

 Diphysis, Megachile ) 

 Social bees \ 



96 9"6 654 I7'3 827 5'i 



5 4 



5'4 54 857 gri 35 

 106 



Bombus apis 



o"5 101 loO 274 437 7fi i3'o 3> I'f) 



o"8 8"3 9'i 263 js"6 8i'9 40 4"8 oa 

 3-1 15s 186 353 345 598 13-4 7-2 i-o 



Explanation of Tabic — 



a, flowers with honey apparent ; b, partially apparent ; c, honay concealed 

 but easily reached ; d, concealed but adapted to a short proboscis ; e, 

 flowers only possessing pollen ; f. useless attempts to get honey by 

 makine a hofe. as on Dielvtra : sr, flowers without honev. 



