Is it possible to make a laser by sending sunlight through an optical apparatus? 1) Is it possible to create a laser from focused sun light by separating and using only one wavelength of light as a laser and using the proper mechanism to polarize it and make it coherent?                                                                                           
2) If so, would it be possible to use some type of a wavelength filter enabling it to focus different wavelengths for different applications?                                                            
 A: If you're asking about the possibility of making coherent light by filtering incoherent, broad-spectrum sunlight, the answer is "no".  You could filter the light down to a very narrow wavelength band by throwing away all the other wavelengths, and end up with nearly monochromatic light. That leaves about 1/10000 of the original light power.  Now, to obtain spatial coherence (which allows a laser to be focused to a small spot), you would need to put a tiny pinhole (a micron or two wide) at the focus of the filtered sunlight.  Try focusing sunlight and you'll find that it is difficult to get a spot smaller than about a millimeter. So the pinhole throws away at least 99.999 % of the monochromatic light.  Now you've got only 1/10,000,000,000 of the light you started with -- and it's still not as coherent as a laser.
On the other hand, if the objective is simply to get highly monochromatic light for experiments such as testing spectral response of photosynthesis or light sensors, all you would need is a prism or diffraction grating, a slit filter, and a lens.
A: Even with perfect filtering (answer of S. McGrew), one can still distinguish filtered sunlight from true laser light by using the 2nd-order autocorrelation function $g^2(τ)$.
For a laser: $g^2(0) = 1$.
For thermal light (e.g. sunlight): $g^2(0) = 2$
In words: While the temporal arrival time of laser photons is randomly distributed (poissonian distribution), thermal photons arrive in small bunches (Bose-Einstein distribution). This can be measured with a Hanbury-Brown-Twiss experiment.
