telescope-spectrograph is used in 1st order with a grating blazed at 4000 . It has a dispersion of 200 per mm on its CCD detector, which is 3000 pixels in the dispersion direction and 2000 pixels in the spatial direction. The CCD pixels are 9.00 m square and the projected 1 arcsec slit width gives 3 pixels per spectral resolution element. a) If the tilt of the grating is set so that the blue (short wavelength) end of an observed spectrum is 3200 , what wavelength will correspond to the red (long wavelength) end of the spectrum?

b) If the seeing improved dramatically from 1.0 arcsec to 0.5 arcsec, how narrow could you make the spectrograph slit and still preserve Nyquist sampling in the dispersion direction?