[R] disturbed legend in ggplot2

[Troels Ring]

Dear friends - below is a subset of a much larger material showing two 
ways of generating two "lines". The intention is to have the colour 
reflect a variable, pH, but the legend is disturbed. The little part 
marked "3" above the colour scale is unwelcome. Why did it appear? How 
could I avoid it?

I'm on Windows 7, R version 3.4.1 (2017-06-30) -- "Single Candle"

All best wishes

Troels Ring
Aalborg, Denmark

library(ggplot2)
DF1 <-
structure(list(P = c(0, 0.00222222222222222, 0.00444444444444444,
0.00666666666666667, 0.00888888888888889, 0.0111111111111111,
0.0133333333333333, 0.0155555555555556, 0.0177777777777778, 0.02,
0, 0.00222222222222222, 0.00444444444444444, 0.00666666666666667,
0.00888888888888889, 0.0111111111111111, 0.0133333333333333,
0.0155555555555556, 0.0177777777777778, 0.02), pH = c(12.3979595548431,
12.3129161148613, 12.2070984076445, 12.0669463736967, 11.8586790792785,
11.4437319273717, 7.64497330556925, 6.98905682614207, 6.63520883742788,
6.3229313658492, 12.176061323132, 12.0234712172719, 11.7861230637902,
11.2219147985144, 7.14240749824074, 6.53119941380901, 5.95522932117427,
3.25184520894594, 2.55614400932465, 2.30097494287507), BC = 
c(0.0576574111386315,
0.047331331067055, 0.037206026657832, 0.0268607893098731, 
0.0166183791472022,
0.00639593998967551, 0.00335972794444094, 0.00854377959176608,
0.00987464693654883, 0.00863636089604445, 0.0343718830720469,
0.0242985554593397, 0.0140710602077036, 0.00383913993097999,
0.00439784065436743, 0.00582135949288444, 0.00336240952299985,
0.00129948001017736, 0.00640073762860721, 0.0115158433720248),
     SID = c(25, 25, 25, 25, 25, 25, 25, 25, 25, 25, 15, 15, 15,
     15, 15, 15, 15, 15, 15, 15)), .Names = c("P", "pH", "BC",
"SID"), row.names = c(NA, -20L), class = "data.frame")

df1 <- subset(DF1,SID==25)
df2 <- subset(DF1,SID==15)
v <- ggplot()
v <- v + geom_line(data=df1, aes(x=P, y=BC,col=pH,size=3))
v1 <- v + geom_line(data=df2, aes(x=P, y=BC,col=pH,size=3))

v <- ggplot()
v <- v + geom_line(data=DF1, aes(x=P, y=BC,group=SID,col=pH,size=3))