332 



TflE AMERICAN 



the other hand, as there certainly is not more than one- 

 hundredth part of the black oaks growing near Rock Island 

 that produce any oak-ajiples at all in any particular year, the 

 chances are at least 99 to 1 against any particular bhick oak 

 there, taken indiscriminately, bearing ouk-apjjles in any 

 particular year. Conycfjuently, the compound chance of any 

 l)LU"ticular Ijhick oak, (aken at random from those growing 

 near Kock Island, not only bearing thon in aparticnlai- year, 

 Ijnt, bearing them cxchisively upon a particular portion df ils 

 bon;,dis,i)reviously designated and forming only onc-twcnlicll] 

 part of the entire mass of its boughs, is, according to i,lie 

 theory of cliances, ojooi oi"? '» ordinary jiarlanoc, the chances 

 arc 1,909 to 1 against such an improl^ii.ble event huppening. 

 Bnt these chances are founded on the supi>usil,ion oL" the tree 

 that was exi)erimented on having been tiiktm at riiiidom with- 

 out any selection; whereas, instead of taking it at riindoni, 

 I exercised the greatest possible care in every instance, 1st, 

 to select a black oak that grew a long way oil' from any other 

 )>lack oaks, ami, 2nd, to select one that I was familiar with 

 and had watched for years, and knew not to have borne any 

 oak-apples for several years preceding. JtisdilUcult, and 

 in fact almost im|iossible, to estimate in ligures to what an 

 extent these additional precautions inci-oascd the odds speci- 

 fied above; but it is <iuite clear that the increase i^iust have 

 been very great. Taking everything into consideration, we 

 may concbnle, with a degree of certainty amountijig to moral 

 conviction, that it was not mere chance that caused the four 

 oak-apples to grow upon the large black oak tluit I was at 

 this period experimenting on, but that they were generated 

 by the gall-ilies 1 had myself placed there in the preceding 

 autumn. And a similar mode of reasoning will apply to the 

 other experiment, the results of which will be subsequently 

 given in detail. 



Some persons, perhaps, who are not familiar with the 

 theory of chances, may consider such odds as those apeciOed 

 just now to be insullicient to produce moral conviction . Yet 

 the very same men, when serving on a jury, will feel no 

 scruple at condemning a prisoner to the gallows for murder, 

 although it is mathematically demonstrable that the chance 

 of any supposed nuirderer taken at random, who is hung for 

 murder, being really guilty of that murder is only about 1999 

 to 1. Scarcely a single case of murder, where the prisoner 

 has been able to cmidoy distinguished counsel, has ever been 

 tried bnt those counsel have enumerated to the jury scores of 

 cases that are on recoi-d, where men that had been tried, 

 convicted, sentenced and actually lumg for murder, were 

 snbsecpiently iiroved to have been entirely innocent of the 

 crime laid to their cliarge. I am satisfied, on a careful ex- 

 amination of the ftxcts, that out of the whole number of men 

 actually hung for murder, somewhere about 1 in every 2000 

 were innocent of the crime for which they suffered. Conse- 

 quently, upon this supposition, the chance of any one sup- 

 posed murderer, taken indisci'immately from the whole mass 

 ol men actually hung for murder, being in reality guilty is 

 0"ly S> oiS i» popular language, the odds are only 1999 to 1 

 that he is guilty. And yet men are every day hung for mur- 

 der when the moral certainty of their guilt is demonstrably 

 very much less than tlic moral certainty of the oak-apples 

 that I was experimenting with having been produced by the 

 gall-llics that I placed upon a particular bough of a particular 

 oak in the pi'eceding autumn ! 



The third black oak, visited on May 21, ISti.!— which was 

 the small one— was absolutely loaded down with galls, and I 

 estimated their number at m or UO at least. At this early 

 period all these galls were still small and immature, and it 

 was necessary to leave them for a week or two upon the tree 

 to ripen and mature. 



On June 5th, lSf?5, I climbed the small gall-bearing black 

 oak, and stripped it of every gall that I could see. From it 

 I harvested only IS normal spongijica galls, exclusive of 2 or 

 3 that had been destroyed by lepidopterous larva3, and about 

 40 specimens of a particular form of gall (Q. pseudolinctoricc, 

 Walsh) which occurs commonly but sparingly among the 



normal Q. spongifica galls, and also, in a slightly modified 

 type, upon red oaks infested by the Q. inanis gall . For a long 

 series of j-^ears this Q. pseudotinctorics gall has been a great 

 puzzle to me; for, whether obtained from black oak or from 

 red oak, although I have bred from hundreds of them, and 

 have kejit them on hand for years, I have invariably bred 

 nothing from them bat great numbers of a very large and 

 very handsome Chalets lly, belonging to the Pteromalidcs^Yii^l 

 ciinceMleil ovipositors, whicli I have never reared from any 

 oilier gall, and a few stray specimens of such C/nr/ris Hies 

 {(Uilliniijine and Eiirytoma) as I have bred also from tln^ uov- 

 nial Q. spongijica galls. My friend liaron Usten Sacken, to 

 whom I had before this period communicated this peculiar 

 form of gall, suggested that it was a true Q. s^pongifica gnU, 

 modilied by the action of the parasite that inhabited it; and 

 tile negative fact that I could never breed anything but para- 

 sites from it, after experimenting with hundreds of speci- 

 mens in three or four different years, compels me to acquiesce 

 in this most anomalous and, so far as I am aware, unprece- 

 dented conclusion. The gall in question never exceeds 0.85 

 inch in diameter, while the normal Q. spongifica gall olten 

 attains a diameter of 1.75 Inch, and is shaped like the normal 

 gall, except that it is often studded outside with sharp 

 prickle-like tubercles similar to those of the exotic gnllcE- 

 thictoricE gall — whence the name that I have given it. The 

 central cell is round and about 0.20 inch in dliuneter, with an 

 external crust which is only about 0.02 inch thick, insteati of 

 forming a dense woody mass as in the normal form. The 

 external crust of the gall itself is similar to that of the nor- 

 mal gall ; but, instead of its being connected with the centi-al 

 cell by homogeneous siiongy matter, with a few subobsolete 

 slender radiating lilaments among it, as in that gall, it is 

 connected with the central cell solelj"" and exclusively by 

 dense, ojiaque, coarse, whitish cottony fibres, radiatingfrom 

 the centi'al cell, as in the Q. inanis gall, but diJlering widely 

 from those of that gall by being very much coarser, by being 

 cottony instead of smooth, and by bciug i)laced so close to- 

 gether as to occupy the whole space between the cell and the 

 extermil crust of the gall, instead of being separated from 

 each other by very wide interspaces. On nay cutting into 24 

 galls that remained unbred from, on this 17th day of March, 

 1809, out of the above-mentioned lot of about 40 Q. pseado- 

 tinctoria; galls harvested June 5th, 1SG5, eight of them were 

 found to contain the dead Pteromalidons imago already 

 spoken of, seven what was probably its mature dead larva, 

 one what was probably its mature dead pupa, one the pujial 

 shell of a Q Callimome, one a Eu^ijtoma. s/udiosa 5 , ^"Yj '"<'l 

 in six (he tenant of the cell must have ]>erished in early life, 

 for in these six the C':'ntral cell wa^ cmjity. 



On the supposition of the peculiar character of the Q. 

 pseudotinctoria:. gall being caused by the action of the large 

 Pteromalidons parasite that generally, but not always, in- 

 habits it, and never inhabits the normal type of gall, it juay 

 be asked how it happens that this very same Q.pseudotvicforicr. 

 gall sometimes proiluces the same species of green (Ja/Umomd 

 which is commonly bred from the normal gall, and occasion- 

 ally a, Eurytoma, which is also bred occasionally from the 

 normal gall? lean only suggest that, in these two latter 

 cases, the Callimome and the Eurytoma arc ])arasitic U|ion 

 the large Pteromalidous parasite, and that tlie peculiar char- 

 acter of the gall was determined in the lirst instauce by the 

 rteromalide. The Chalcididcc are, to a much greater extent 

 than is commonly supposed, secondar-y and not in-imary 

 liarasites; and in the case ol the Joint-worm Fly (Isosoina 

 hordei, Harris) we have an instance of a Eurytomiduus Chal- 

 cidian being jireyed upon parasitically by three otlier Chal- 

 cidians— one a TorymuSj a genus closely allied to Callimome, 

 and the other two belonging to the Pleromalides with short 

 ovipositors. 



From the 18 normal Q. spongifica galls, obtained on June 

 5th, ISCiH, as specified above, from the small black oak, I bred 

 on June 11th, 1S65, one C. q. sp07igijlca 9, and another $ of 

 the same type on June 14th, 18G5. On cutting into the re- 



