Merge pull request #9 from tschmidt23/master

a few spelling corrections

doc/cramer.h

@@ -5,7 +5,7 @@ TooN aims to be a fast library, and may choose between one of several algorithms
 depending on the size of the arguments. However TooN will never substitute a
 fast but numerically inferior algorithm.  For example LU decomposition, Gaussian
 elimination and Gauss-Jordan reduction all have similar numerical properties for
-computing a matrix inverse. Direct inversino using Cramer's rule is
+computing a matrix inverse. Direct inversion using Cramer's rule is
 significantly less stable, even for 3x3 matrices.
 
 The following code computes a matrix inverse of the ill conditioned matrix:

doc/documentation.h

@@ -216,7 +216,7 @@ This section is arranged as a FAQ. Most answers include code fragments. Assume
 			 double* my_data;
 		@endcode
 		When a Vector is made <code>const</code>, C++ inserts <code>const</code> in
-		to those types.  The <code>const</code> it inserts it top level, so these
+		to those types.  The <code>const</code> it inserts is top level, so these
 		become (respectively):
 		@code
 			 const double my_data[3];
@@ -363,7 +363,7 @@ This section is arranged as a FAQ. Most answers include code fragments. Assume
 			Matrix<2> m = Idendity;
 			Matrix<> m2 = Identity(3);
 		@endcode
-		note that you need to specify the size in the dynamic case.
+		Note that you need to specify the size in the dynamic case.
 
 		Matrices can be filled from data in row-major order:
 		@code