408 



Single-stem trainhifj (pp. int-l .">(;).— ••In the account of forcing toma- 

 toes, given iu Bulletin No. 28 of the station [see Experiment Station 

 Eecord, vol. iii, p. 01]. it was stated that the yield was about 2 pounds 

 per square foot of floor space, which is over three times the yield of a 

 good crop of outdoor tomatoes." To determine whether the house sys- 

 tem of training is i)racticable for field culture several experiments were 

 made. 



We used iu oue experiment stakes 1^ by li inches and 5 feet high, driven securely 

 into the ground. In another we drove stakes in lightly and held them iu place by 

 two wires drawn in between them iu an alternating fashion near their tops, the wires 

 being held by posts sot every 16 feet. This made a wire and picket fence, except that 

 the lower ends of the pickets were held iu the grouufl. This is the cheapest and 

 best trellis which we tried. » * * 



We also used perpendicular strings stretched lictweeu two horizontal wires, one 

 near the <'round and the other 5 feet high, and also one or two other trellises, but 

 tlu-v were not so good as stakes supjiorted by wire. The phint was tied loosely to 

 the support, at intervals of a foot or so, as it gr4-w, and all side shoots w«ii-iiinclied 

 out as fast as they appeared. In this way only one stiniwas allowed to grow. Tlie 

 plants were set a foot ai>art in the row, and the rows were 3 feet apart. Several 

 varieties were tried, but only the average results are given. * » » 



Two series of trellises were erected, and the s:iiiie number of plants was set in an 

 ordinary i»atch alongside each one. » • » 



Th«' trellis plants gave 1.6 pounds of rii)e fruit to each square foot, while the other 

 plants gave 0.7.5 ponud, or less than half as much. It is to be uoted also that the 

 former gave earlier results. We also found that rot was much less upon the trellis 

 plants than upon the others. 



It is doubtful if this single-stem training c.in be jtrufitabiy used tur the main lield 

 croi)s, but for early market or choice tradi- aiul Ibr iiome use it appears to possess 

 decided ailvantag<-s. 



JlilliiKi (]K l"»i>). — An cxpcriiiiciit witli liiiintiiiii i»l;iiils on pom' chiy 

 loam showed no ad\ ;nil;i,i;<' t'mni liillinu as rc,i;;ii(ls cither earliness (»r 

 productiveness. 



"iy<^/7////"7>/"»^">"(pp- 1 •'»*'». '•'">• — IJuryin."; the stems of badly drawn ]»lants 

 in a Ixtri/ontal position so :is to lea\e only 10 or 12 inches of the stem erect, 

 increased earliness. nnmlicr of fruits ]»er phint. ;ind weight of «'i-op. Nor- 

 ni;il i»l;ints, ]i»»we\c!-. naxc nunh better results than " leu'uy" plants, 

 even wlien the liilter weic set deeji. 



rnidiK-ts of airhf a ml Inlc I'lidts (pp. |.'>7- 1 ."■•!•). — 'I'he results ;ire tabu- 

 lated of an e\i>eiiMient in whi<'li seeds of S varieties from IVuits |>icked 

 August "» and Sei»t ember 17. were compared. The averages for the seeds 

 of the resperti\c <hites wfre as tbllows: Numlur of fruits per jilant 30.7 

 and ■■)<).!>, weight of fruits jjcr ]»lant 7.."> and !>. 1 pounds, weight of indi- 

 vidual fruits 4.1 and 4.(! ounces. In e;iiliness tln-re was no ('(Uistaut 

 ditference. Kererence is ininie to experiments reported in Uulb'tin No. 

 4Softhe Michigan Statiiui (see lOxpcrinu'Ut Statittu IJcccU'd, vid. 1. ]>. S!») 

 in which the curly anguhir varieties gave better results fr(Mn seed from 

 lirst fruits, while the reverse was true of the round or apple shaju'd 

 varieties. The experiments thus far do not warrant detinite con<'lusions, 

 but "little if anything appears to be gained by selecting seeds from lirst 



