NOTES ON LUFFIAS. 151 



no doubt be admitted, but when there is really no evidence of the 

 slightest character in support of such a series of suppositions, whilst 

 the other theory has many facts that point conclusively to it, I would, 

 if I might be permitted to do so without disrespect, give my comment 

 on the passage in the form of a common Scotch saying, to suggest 

 improbability, " It might happen a cat might kittle a piet." 



The saAvflies are another family of hymenoptera that give us another 

 phase of parthenogenesis, and one that seems to be nearest to that 

 we meet with in lepidoptera. There are a good number of sawflies 

 that are parthenogenetic, and of which yet males occur with greater or 

 less rarity, whilst there are at least a few in which it seems probable 

 that not only are males not met with because they are very few and 

 rare, but really because there are none. 



In the lepidoptera parthenogenesis occurs in what we may call a 

 sporadic manner, i.e., occasionally, unfertilised eggs, which would 

 usually undergo no development whatever, do develop, more or less, 

 and, as a very rare occurrence indeed, proceed so far as to hatch and 

 produce veritable larvae. It is very probable that sporadic partheno- 

 genesis of this character would be found to be equally common in 

 other orders, were rearing them from the egg as frequently practised 

 as in lepidoptera. It is this sporadic parthenogenesis that no doubt 

 affords the initial material from which the forms we have already 

 noticed are developed by natural selection under suitable conditions. 

 This sporadic j^arthenogenesis gives rise to progeny of both sexes, but 

 in a good many instances, I think, very decidedly in the majority of 

 cases, the strong tendency is to produce males, and instances are 

 recorded where experiments in continuing, or attempting to continue, 

 a sporadic parthenogenetic variety came to an end because only males 

 Ave re produced. 



Taking these cases in connection with those in hymenoptera, we 

 may surmise that the male influence is to produce female offspring 

 and the female to produce male offspring. If this be so, Ave may hide 

 our ignorance of hoAv it came about, by attributing it to innate condi- 

 tions, arising along Avith the distinction into tAvo sexes. But, I think, 

 AA'e may get a little nearer a comprehensible explanation, if Ave assume 

 that naturally both sexes are produced equally, but that variation may, 

 and, frequently did, produce races that abound more in one sex than 

 the other. If this Avent too far, so as to injure the prospects of the race, 

 a remedy Avould be beneficial, this Avould accrue if variation arose by 

 Avhich each sex tended to produce the other, and so an excess of 

 females Avould tend to produce more males and of males to produce 

 females. It arises here to ask hoAv could this effect any result Avithout 

 parthenogenesis, not only in females but also in males. There seems, 

 hoAA-ever, little diflticulty in the matter ; the fact that parthenogenesis 

 occurs, shows that very varying proportions of the male and female 

 elements in an egg, may equally conduce to its proA'ing fertile, even to 

 the extent of its possessing no male element at all. The process of 

 fertilisation of the eggs, in most insects, shoAvs that such A'ariations 

 may A^ery easily occur. In lepidoptera both sexes frequently pan- 

 more than once. If males Avcre scarce, each male Avould pair Avith 

 seA'eral females and AA'Ould supply the spermathecie Avith but a scanty 

 content, and each ovum Avould receive the minimum supply of the 

 male element, at any rate, in all but the first mating of the male. On 



