SCIENCE-GOSSIP. 



209 



" Butterflies and Moths of Teneriffe," also includes 

 Aporia cratagi, Thecla rubi and Danais dorippus, 

 var. klugii, but the occurrence of, at any rate, the 

 former two, is, I think, extremely problematical. 



On looking at the above list, one is at once struck 

 by the fact that in these sub-tropical islands, lying 

 near the African coast, the butterflies are almost 

 entirely of North European type. If we except 

 Euchloe charlonia, a desert species, which is reported 

 to have been found in the eastern islands, and 

 probably wind-borne from the Sahara, no typical 

 African species occur. Not taking into considera- 

 tion the endemic species, Lyccena webbiana, there are 

 seventeen species which are found in North Europe, 

 and five which have their prototypes there, Pieris 

 cheiranthi, Epiphenele hispulla, Pararge xiphioides, 

 Gonepteryx cleobide, and Lyccena cestiva Indeed, it 

 might be considered by some that these are only 

 modifications of the species, due to climatic effect. 

 Follow, for instance, the type of Gonepteryx rhamni. 

 In North Europe the orange spot on the forewing 

 of the male is quite small, increasing in size as it is 

 found further south, until in Madeira (in G. cleo- 

 patra) it covers more than half the front wing, 

 whilst still further south, in the Canary Islands, 

 the entire front wing is diffused with it. In tro- 

 pical Africa it is replaced by Callidryas florella. 



The butterflies of the Canary Islands appear, 

 then, to have a direct relationship with those of 

 North Europe, and not those of West Africa, as 

 might be expected. It is a pity that more is not 

 known of the butterflies of Morocco, the examina- 

 tion of which might throw some light on this 

 interesting subject. 



Sir James Hooker and Mr. Ball, in their excel- 

 lent work on Morocco, " Morocco and the Great 

 Atlas," appear to have left unnoticed nothing but 

 the insects of that country. 



Santa Cruz, Teneriffe ; September, 1895. 



VARIABILITY OF ELDER- 

 FLOWERS. 



By C. E. Bretton. 



T F an inflorescence of the common elder is 

 carefully and closely examined, it will be 

 noticed that some of the flowers differ from the 

 structure of the typical elder-flower. There is a 

 tendency to lessen the number of members com- 

 posing each floral whorl, so that in place of 

 pentamerous flowers, we notice flowers with the 

 parts arranged in fours. An accepted theory, 

 accounting for the occurrence of the tetramerous 

 flowers, ascribes their origin to insufficient supplies 

 of nutrient matters. Two inflorescences, taken 

 from different trees growing close together, were 

 carefully examined by the aid of a lens, and the 

 number of flowers differing from the type, and the 

 particulars in which they differed, were noted. 



Previous to this, it. was found by counting the 

 number of flowers of several corymbs, that a low 

 average number of flowers in an inflorescence was 

 400. The first inflorescence, A , taken from one tree, 

 showed thirty-six flowers deviating from the type. 

 These were distributed as follows : Sixteen of the 

 thirty-six showed a reduction of one in the number 

 of members constituting the three outer whorls of 

 the flower, the pistil being unaffected. If we put 

 down the formula of the typical flower as S5 P5 

 St5 C3 = sepals 5, petals 5, stamens 5, carpels 3, we 

 may contrast with this the formula of the sixteen 

 non-typical flowers, S4 P4 St4 C3. Five flowers 

 showed that a reduction by one member had 

 affected all the whorls. The formula of these 

 flowers would be S4 P4 St4 C2. Another series of 

 five flowers showed that the reduction had affected 

 the innermost whorl, the carpels only. We may 

 denote the structure of these flowers by S5 P5 St5 

 C2. So far we have accounted for 26 non-typical 

 flowers. Of the remaining ten, three showed a 

 decrease in the number of members of the floral 

 whorls, whilst seven showed an increase. Taking 

 those showing a reduction first, we have flowers 

 with the structure— (1) S4 P5 St4 C3, (2) S5 P4 

 St4 C3, (3) S5 P5 St4 C3. Two of the seven 

 showing an increase in number of members of the 

 whorls, had the structure expressed by the formula 



55 P5 St5 C4. Two more, instead of having the 

 carpels increased in number, had the other mem- 

 bers increased, the formula being S6 P6 St6 C3. 

 The following expressions denote the structure of 

 the remaining flowers— S5 P6 St6 C3, S6 P6 

 St6 C4, S7 P8 St8 C7. The last formula is rather 

 curious, and the abnormal flower may have been 

 due to two growing-points of flowers arising in the 

 place of one, or to the union of two originally 

 distinct growing-points. 



In the second inflorescence, B, taken from the 

 other tree, more than one-fourth of the entire 

 number of flowers were non-typical. To be precise, 

 the exact number was 108. Whereas in the first 

 inflorescence, A, the chief form of non-typical 

 flower had the structure S4 P4 St4 C3, in B this 

 type was represented only by seven flowers, as 

 contrasted with seventeen in A . Again, in B the 

 prevailing form of non-typical, amounting to sixty- 

 seven of the total 108, was S5 P5 St5 C2. In A only 

 five flowers were constructed on this plan. Also in A 

 five flowers had the formula S4 P4 St4 C2 ; in B 

 twenty-eight flowers were of this type. One of the 

 remaining six flowers showed a great increase in 

 the number of its parts, and was probably due to 

 the same causes that produced the similar flower 

 in A . This and the other flowers had the structure 

 set forth in the following formulae : S8 P8 Stg C8, 



56 P6 St6 C3 (two flowers), S6 P6 St6 C2, 

 S5 P6 St6 C2, S5 P5 St 4 C2. 



189, Beresford Street, CamberwelL S.E.; Sept.. 1895. 



