In his 1916 General Theory of Relativity, Einstein predicted the existence of gravitational waves. These can be considered as ripples or waves in the curvature of space-time. Until now there has only been indirect evidence for their existence, and scientists around the world are working on a number of gravitational wave detectors. The direct detection of these waves will provide information about astrophysical processes and the sources that produce them. Gravitational waves are quadropole in nature, producing a tidal strain in space. Their interaction with matter, though, is very weak, making them very hard to detect. Gravitational waves emitted by violent astrophysical events are predicted to produce strains at the Earth of the order of 10e-21 to 10e-22 at frequencies accessible to ground-based detectors (~10 Hz to a few kHz). One way to detect these strains is based on the Michelson interferometer. The Institute for Gravitational Research at the University of Glasgow, led by Prof. J. Hough, is working with collaborators from the Max-Planck-Institut fur Quantenoptik at Hannover and Garching, the University of Hannover, the University of Cardiff and the Albert-Einstein-Institut at Golm on a project called GEO 600 to build a laser interferometer with arms of length 600 m. The GEO 600 detector is close to completion in Germany. GEO 600 is designed to operate down to 50 Hz. The sensitivity limit at this frequency is set by the thermal noise from the internal modes of the test masses that form the interferometer mirrors. The level of this noise is determined by the level of mechanical loss factors within the system. For GEO 600 these are constructed from fused silica. The strain sensitivity limit from thermal noise is expected to be 2 x 10e-22/√Hz at 50 Hz. To improve sensitivity, the GEO 600 technology will need to be transplanted to a system of longer arm length, such as the American 4 km detector LIGO. Beyond this, any further increase in the sensitivity further will require a different material with a lower loss than fused silica to be found. The work contained in this thesis covers the testing of the mechanical loss factors of a range of test substrates to determine their potential use in inter-ferometric gravitational wave detectors. In GEO 600 and all other detectors currently under construction, fused silica is being used for the test mirrors due to its high intrinsic quality factors (i.e. low loss factors). Samples of fused silica were tested and the good loss values confirmed. Different types of silica, though, resulted in a range of quality fac-tors, highlighting the need for good manufacturing if the low loss levels were to be maintained. The intrinsic loss of fused silica currently limits the sensitivity of GEO 600, and if this limit is to be improved upon a different substrate with a higher intrinsic Q will be needed. A range of suitable contenders to replace fused silica was tested, the best results being achieved with sapphire samples. It had long been known that sapphire had an intrinsic Q almost 10 times that of the best fused silica samples, but this had not been achieved for samples tested outside of Russia. Work carried out on HEMEX samples yielded one example with a Q of 2.59 x 10e8, the highest intrinsic Q measured outside of Russia. This confirmed it as the front-runner to replace fused silica. (Though potential problems resulting from its relatively high thermal expansion coefficient make it less attractive.) In order that the fused silica test masses may function as mirrors in the in-terferometric detectors, they are coated with dielectric mirror coatings. To discover whether the application of these coatings increased the losses of the test masses to a point where the sensitivity of the detectors would be degraded, a range of resonant modes for two coated fused silica masses were investigated. Analysis of the varying Q results from these tests, along with FEA modelling of the modes led to the conclusion that the coating would not present a problem for the current generation of detectors, but may prove significant for future detectors, such as Advanced LIGO. Tests carried out on a coated sample of sapphire were suggestive of a similar situation existing. In order that the jointing of test mass to suspension system in GEO 600 will not contribute to the loss levels of the mass, they are suspended from fused silica fibres welded to ears that have been hydroxide catalysis bonded to the sides of the mass. It was already known that this bonding process did not affect the loss level significantly for fused silica to fused silica bonding. However it was not known whether this was true when fused silica was bonded to sapphire. A fused silica post was therefore bonded to a sapphire sample. Results suggested a drop in Q of roughly a factor of 10. The post had been bonded to the mass by a variation of hydroxide catalysis bonding where a sodium silicate solution was used in place of the traditional potassium hydroxide solution. This was done because this type of bonding appeared to result in a stronger bond. Results here suggested that the bonding with sodium silicate solution resulted in excess losses due to the bond of a similar order to those for the KOH case, when scaled to the case for GEO 600.